com
The concepts and formulation of quantum mechanics are not elementary in the
sense that they are easily understood. They are based on the outcome of considerable
theoretical research supported by experimental evidence. The quantum mechanical
approach to physical problems cannot be explained in simple non- mathematical terms.
Simple and elegant mathematical techniques have been used to elucidate the physical
concepts.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
During the decades of 20th century, Physicists turned their attention to the
study of atomic systems, which are inherently inaccessible to direct observation. It
soon became clear that the concepts and methods of classical macroscopic physics
could not be applied directly to atomic phenomena.
associated with the physical entity, the quantum state, is known as the State Vector.
The different kinds of wave functions for a given state (in the configuration space,
momentum space) are simply the sets of components of the state vector as referred to
different coordinate systems or bases in the state vector space. This notation was
introduced by Dirac. It is used to relate different representations. The notation is very
general and suitable for handling of systems which have no classical counterpart, e.g.
spin angular momentum.
Bra-ket notation is the standard notation for describing quantum states in the
theory of quantum mechanics. It can also be used to denote abstract vectors and
linear functionals in pure mathematics. It is so called because the inner product (or
dot product) of two states is denoted by a bracket, consisting of a left part, called the
Bra, and a right part, called the Ket. The notation was invented by Paul Dirac, and is
also known as Dirac notation.
from its representation, the wave functions. The state vectors constitute a complex
vectors space, i.e. if y , c are state vectors, so that c y +c 'c for any complex
numbers c and c'. all vectors c y , c 0, which have the same direction y ,
correspond to the same physical state. The set of all such vectors constitute a space
This watermark does not appear in the registered version - http://www.clicktoconvert.com
called a ray in the space. Therefore, we can say that any physical state is represented
by a ray (rather than a single vector) in the vector space.
i.e. y y , y y (1)
The relation between any vector and its conjugate may be visualized readily if
we consider examples of their representations. In the Schrdinger picture
y y(x y x (*
) , y ). It is to be mentioned that the relation between y and
y is not a linear.
i.e. y y 0y ,y
= 0 if and only if y = 0 . (2a)
In general if the vector y and its conjugate f of some vector f are juxtaposed in
such a was as to form a closed bracket f y , this symbol represents a single (real or
that order. Vectors of type y which forms the first half of the closed brackets are
called Bra Vectors and those of the type y which forms the second half are called
Ket Vectors. The conjugate of a bra vector is a ket vector, and vice versa.
Any vector space (in general infinite dimensional) such that between any two
vectors of the space a scalar product which satisfies equation (2) is called a Hibert
space. The space of ket vectors (or bra vectors) representing quantum mechanical
states is thus a Hilbert space.
dynamic variables in quantum mechanics do not commute with each other that is
AB BA.
( e , ( ) e,= d1
i j
, i 2= , j. . . N. , ) ,
i j (1)
N
ei =
uj *j
e
i j (2)
j =1
in which ui j are suitably chosen complex numbers. This new basis are also
orthonormal, provided that
(e , e ) = d
i' j'
i j (3)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
*
u
k
u =d
i k j k i j (5)
U U = 1
U -1 =U (6)
It is also apparent that this conditions is sufficient to insure that the vectors ei ' formed
according to equation (2) are orthonormal. Hence the necessary and sufficient
conditions for the vectors ei ' to form an orthonormal basis is that the matrix U is
unitary. Since U -1U =U-1 U= 1 , the equation (6) becomes
U U = U U
= 1 (7)
Such a unitary transformation has the effect of substituting the basis {e' } for the basis
x=
x i ei (8)
i
' i'
x=
x i e (9)
i
Then, since these expression representatives of one and the same vector,
j
x i i
e= xi ju' j * e (10)
j i j,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
hence by orthogonality of e j ,
u j * xi '
x j = i (11)
i
This equation provides the connection between the components xi ' xi of the
vector with respect to the bases {e} {e' } . Since U is the unitary matrix, equation (11)
1.12. References
1.A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan. Tata
McGraw Hill Publishers.
2.Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
3.Quantum Mechanics V. Devanathan Narosa Publishing house.
4.Quantum Mechanics S.L. Gupta, V. Kumar, H.V. Sharma, R.C. Sharma. Jai
Prakash Nath & Co Ltd.
5.Quantum Mechanics Satya Prakash Sultan Chand Publishers .
This watermark does not appear in the registered version - http://www.clicktoconvert.com
LESSON - 2
EQUATION OF MOTION
10
d
-i h at s =( ) tas ( H)
d t
= a s (t ) H (2)
This can be easily verified by differentiating equation (3) and getting the equation (1)
-i H t
d i -
a s ( t ) = H as h
e (t )
d t h
d
ih at s ( ) =as
H t( )
d t
-i H t
h
The operator e is an infinite sum of powers of H , each of which is a
dynamical variable, that can be expressed as an operator or square matrix. It is
-i H t
h
essential to remember that the operator e is not an unitary transformation in the
ordinary sense of rotation of axes in Hilberts space as U, V, W. But it represents the
change of in to some other ket a s ( 0 ) . Thus the series as a whole is also a dynamical
variable. The time rate of change of matrix element of a dynamical variable W s in the
Schrdinger picture can be found out as
d d dW
a s (t W) s b (t s =) a{ s(t }W
) b (t
s s ) +a t(s ) b+ss t( s) aW t( s ) b s t{( ) }
d t d t t d t
-1 1 W s
= a s (t W) s bH ( s )+t a (s t ) +bs t( s ) at W( s H) b s ( t )
ih t h i
W 1
= a s (t ) s+s bt ( )s t aW
( s ) H[ b, s ](t )
t i h
The first term on the right hand side arises due to any explicit time dependence
W s may have. The second term on the right hand side arises due to the change in the
state vectors with time.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
11
d W 1
a s (t W) s b t( s )= at( s ) s t+bs( ) t s aW
( )[
H
s b
, s ] (t )
d t t i h
-i H t i H t
with a s (t =) eas (h 0 ) and a s (t )=as ( e0 ) h
i H -t i H t i - H t
d h h h
h i W
s
H t
as ( e0 W s) b (e =0s )ae ( s 0 e )b s ( 0 )
d t t
i H t i H - t
1
( e0h s- H
+ a sW ) (b
W
H
s
h
) e (s 0 )
ih
-i H t
h
Assuming H to commute with e and also choosing the state vectors in
Heisenberg picture to be time independent as
i H t
aH ( 0= as) es h )
(= 0a (t )
since a H (t ) does not depend on time, the explicit mention of t may be omitted. The
dynamical variable W H and its derivative to be
i H T i -H T
h h
W H =e s We and
i H T i -H T
W W
H
=e h esh
t t
i H t - i H t i -H t i H t
d W
H 1 h h h h
aH W H b = H +
aH HW b e H e- Has HW s e b eH
d t t i h
This watermark does not appear in the registered version - http://www.clicktoconvert.com
12
d W 1
a H W Hb = a H H + bH
H HWa [H ,bHH]
d t t i h
d
since the state vectors in Heisenbergs picture are independent of time can be
d t
brought inside the ket a H in the above equations and we will have
d WH H W 1
aH b
H =Ha H+ b H Wa H [ H ]
,bH
d t t i h
This equation is true for arbitrary bra and ket vectors and hence we can write
d W H HW 1
= + W[ H ,H]
d t t ih
d F F
= + { F , H}
d t t
We know that the state vectors alone or the dynamical variable alone are not
central to describe the development of a physical system with time. Both the state
vectors and the dynamical variables can be altered by means of the unitary
i oH s t
h
transformation operator e .
The quantities that are not altered are the matrix element of the dynamical
variable and this provides the physical content of the system since eigenvalues,
expectation values, and transition probabilities can be obtained from the matrix
element.
13
structure as the coloumb interaction and H' may be the external electromagnetic field
varying with time. In the interaction picture we define the state vectors and the
dynamical variables as
- i oH s t
h
a I (t =) a
e ( s t ) (1)
i oH st io- Hs t
W I (t =) es h W eh
i H t i H t
d i o s
do s
ih at I ( ) i= h oH a hes + h (t ) sh ai s e (t )
d t h d t
i oH s t
= H o -a ts( )+ Ie a H h
s s (t )
i oH s t
= H o -a ts( )+ IH a e s
h
s (t )
= H o -a s t( )+I H
as I t( )
= H ' a I (t )
i H t- i H t i H- t i H t -i H t i Hs t
d WI i o s o s
o i W s o s o
s ho
= Ho Wh e s +he s e+h s s e Wh eo -h H s e
d h t t h
W 1
= I
+ ( I Ho- s Ho Ws I )
W
t ih
W 1
= I
+ [ I , Ho s]
W
t ih
W 1
= I
+ [ I , Ho I]
W
t ih
14
2.7. Problems:
Show that If the Hamiltonain H of a system does not contain
time explicitly
i H t
h
a (t ) = e
Solution
t
d
a ( t ) = a(t0 )+ add '' t ( )t '
0
t
1
= a ( 0 )+ a 'H t( )d' t
ih 0
t t
1 d
= a ( 0) + ' aH ' +t( ) t'a d ' (t ) ' d 't
ih 0 d 0 t ' '
t'
1 1
= a ( 0) ( )d ' t
+ 0 H a t'+ d' t ' H ( ' t)
a
th ih 0
1 1 2 2
= 1 +H. t .+ H2 . +t . a0( )
ih 2ih ! ( )
- i H t
h
=e a (0)
- i H t
h
since H is hermitian U is unitary, where U is e
2.8. Do it yourself
This watermark does not appear in the registered version - http://www.clicktoconvert.com
15
2.9. References
1. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan. Tata
McGraw Hill Publishers.
2. Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
3. Quantum Mechanics V. Devanathan Narosa Publishing house.
LESSON 3
MATRIX MECHANICS
16
1
= {ihp} + {ihp}
2m
2ihp
=
2m
ihp
= (2)
m
The commutator of p with H
p 2 1
[ p, ]H= , p + p , k x2
2m 2
1
=0+ k { x[ p, ]x+ [ p
, x] }x since [p,p] = 0
2
1
= k {- ihx - ihx}
2
-2ih k x
=
2
= - ihkx (3)
If there are transition from the initial state l to the final state k
ih p
k [ ,x ]H = l k l
m
ih
= k p l (4)
m
But k [x , H] l =k x H -H x l= k x H
- l k H x l
Introducing a dummy operators, l l and k k
= k x l -l H l k H k k x l
= ( El - Ek ) k x l (5)
From equations (4) and (5)
ih
El - Ek k= x =l
k p l
m
ih k p l
k x l= (6)
m lE- kE
Similarly
k [ p , H] l =k i- kh x= l h-i k k x l (7)
But
k [p , H ] l =k p H -H p l= k p H - l k H p l
= ( El - Ek ) k< | p >| l (8)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
17
-i h k h i k p l
k p l=
( El - E k ) m -(l E k E )
h2k 1
= 2
k p l
m ( El - Ek )
h 2w 2
k p= l k 2
p l
( El - Ek )
( El - Ek =) hw
Thus we see that the energy difference between successive levels will be hw .
Therefore the energy change can takes place in units of hw
(10) + (11)
hw
k p- i w m =x l k w-p i m x l
( El - Ek)
( El - Ek kh-
w) p i -wm 0x = l
18
( p +i w )(
m x -p wi m )=x0 0
{p 2
+i mw2( x - p 2 pw
) +x 2m x =0 } 0
Dividing throughout by 2m
p 2 i 2wm 2 mw 2
x
+ ih + =0 0
2m 2m 2m
p2 1 1 2
- hw + mw x 2= 0 0
2m 2 2
p2 1 1
+ m w 2 x-2 wh = 0 0
2m 2 2
1
H - hw =0 0
2
1
H 0 = h0w
2
1
Thus hw is the eigenvalue of the operator H, corresponding to the eigenket 0 .
2
We can also similarly prove that ( p + i w
m )x increases the eigenvalue by hw . Since
this is the lowest eigenvalue and each operation of ( p + imwx ) increases the eigen
1
En = n + hw n=0,1,2,..
2
( p + imwx ) is called the raising operator or creating operator and ( p - imwx ) is called
the lowering operator or annihilation operator for the above mentioned reasons.
19
1
a = m x {- wi p }
2mhw
1
a = m x {+ wi p }
2mhw
1
a
a= m x{ +wi p m}w{- x i p}
2 mh w
1
=
2 mh w
{ {
m 2wx i+2wm p(-2 x x +2)p p }
(
1 1 2 2 i wm- hi) p 2
= mw x + +
hw 2 2 m 2 m
1 1
= H + hw
hw 2
H 1
= +
hw 2
1
hw a -haw = H (12)
2
1
a a= m x{ +wi p m}w{- x i p}
2 mh w
1
=
2 m hw
{
m 2wx i +2 w
m x(-2 p p) +2 x p }
1 1 2 2 i wm p2
= m +w x +
hw 2 2 m 2 m
1 1
= -
(H h) w
hw 2
H 1
= -
hw 2
1
H = haw +a hw (13)
2
Post multiply by a in (12)
1
H =ahw a - ahw a a (14)
2
Premultiply by a in(13)
1
a =Hhw a + ahw a a (15)
2
Since aH - Ha = [a, H ] we have
[a, H ] = hwa
Similarly
This watermark does not appear in the registered version - http://www.clicktoconvert.com
20
1
a H= hwa -a hwa a
2
1
H =a hwa +ahwa a
2
Since a H- H =a a , H we have
a , H= h-wa
a, a =a a- a a
H 1
a a= +
hw 2
H 1
a a= -
hw 2
a, a =1
This is called the number operator
Eigen values of raising and lowering operator a, a
Let a y n =an yn
a y n =bn yn +1
Let us find
H 1
y n a ayn = yn - yn
hw 2
(n +1 / h2w ) 1
= yn - yn
hw 2
= n y n yn
=n (16)
y n a yn a =yna ayn
= a n+ an yn-1 yn-1
2
= an (17)
Comparing equations (16) and (17)
an = n
a y n =n yn -1
In a similar manner let us find
H 1
y n a ayn = yn + yn
hw 2
H 1
= y n + yn
hw 2
= yn
( n + 1 +/ )y2hw 1
n
hw 2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
21
= ( n +1) y n yn
= n+1 (18)
y n a yn a =yna ayn
= b n bn y
+1n a y
+
n1
2
= bn (19)
Comparing equations (18) and (19)
bn = n +1
a y n =n y+n 1 +1
= nd n' , -
n1
0 2 0 0 0
0 0 3 0 0
a =
0 0 0 4 0
0 0 0 0 5
y n' a yn =n yn +' 1 yn +1
= n + 1d n , +
n1
0 0 0 0
2 0 0 0
=
0 3 0 0
0 0 4 0
1
a = m x (- wi p)
2mhw
1
a = m x (+ wi p)
2mhw
1
a + a= m (2 w x )
2mhw
2mw
= x
h
This watermark does not appear in the registered version - http://www.clicktoconvert.com
22
h
x= a + a
2mw
{ }
0 2 0 0 0 0 0 0 0
h 0 0 3 0 0 2 0 0 0
= +
2mw 0 0
0 4 0 0 3
0 0
0
0 0 0 5 0 0 4 0
0 2 0 0
h 2 0 3 0
=
2mw 0 3 0 0
0 0 4 0
1
Similarly (a
- )a = i -( p2
2hw m
)
mhw
Or p = i - a
2
(-a)
0 -2 0 0
mhw 2 0 -3 0
=i
2 0 3 0 -4
0 0 4 0
When the array of potential wells extends to infinity on both sides, we have to
make sure that the wave functions does not diverge as x .We know that this is
possible only if m = 1 For, ,N
. if m < 1 m and m - N as N +
as N - .
When >1 the situation is reversed. In either case, the wave function in the Nth well
and N -
will diverge in one or both of the limits N + . Hence we must have
Also we have
1
( )
c oq =s mc o m s-1+ha ' .w c =bo s w. s as+i nw h 'b=.f sw i nb -c wo s j ( )
2
(2)
where
This watermark does not appear in the registered version - http://www.clicktoconvert.com
23
1
2 2 2 2
f =( c oa s+s h w' . sai n wh ' ) (3)
t aj n= s. at a w n h '
The above equation (3) determines the allowed energy levels since and
(and hence f and also) depend on the parameter E. From the structure of this
equation it is immediately evident that the allowed energies form continuous bands
with gaps in between. For, since the value of the left hand side of equation (2) is
restricted to -1 c oq s 1 the
, equation can be satisfied only when the right hand side
also is within this range. As the parameter E is varied , the right hand side,
f c ob s w-( j) as, oscillates between the bounds + f (E) and - f (E) which lie outside
for which f c ob sw
- ( j ) 1constitute the allowed energy bands, and those for
which f c ob sw
- ( j >) 1are the forbidden regions. These form energy gaps,
between allowed bands. The variation of f c ob sw- ( j )with E, and the identification
of energy bands and gaps are illustrated in Figure 3.1
V =0
V = -V0
E ' b E
-.25 -.25 V0
V0
-.75 -.75
-2 -1 0 1 2 -2 -1 0 1 2
24
The nature of the energy eigenfunctions u(x) corresponding to any energy E within an
allowed band can now be easily seen. We recall that we have already imposed the
necessary continuity conditions on the wave function, which resulted in the transfer
equation
( ) An+
An-
=Mn
A-
A+
n
( ) -1
-1
(4)
and also the condition of finiteness at infinity which gave us the restriction equation
(2) on the eigenvalues of the transfer matrix M. There are no further conditions on the
wave function, and in particular, there is nothing to prevent us from choosing the two
coefficients (A+, A- or B+, B-) in any one of the regions in Figure.1.2. Suppose then
that we choose some energy E within an allowed band, and after finding ,,c,s for
this E using
2m 2m 2m 2
2
= a 2 a-2 n (+d 2 E0)= bV 0 (-V )=
E (5)
h h h
determine from equation (2) and the matrix M itself from equation (4). Now let us
avail ourselves of the freedom of choosing the coefficients, and take Ae+ , Ae- to be such
that the column made up of these is the eigenvector belonging to the eigenvalue eiq of
M, i.e.
M ( ) e= ( )
A0+
A0-
q
A0+iq
A0-
q
(6)
( )
A+N
AN-
= MA- N
A+
( ) 0
0
(7)
we have
This watermark does not appear in the registered version - http://www.clicktoconvert.com
25
( )
An+
An-
q
A+
= eAi- qn ( )
q
0
0
(8)
i.e. the coefficients depend on n only through the simple factor ein . It is obvious that
a similar result holds good for Bn+ a, n n-d Balso. If we think of a well and the adjacent
barrier together, as forming a unit cell of the repeating pattern, the result just
obtained may be stated in the following instructive form: The wave functions at
corresponding points x and x + nb of wells (or barriers) separated by n unit cells differ
only by the constant factor ei qn :
bq (ei qn u ) x
uq ( x + n) = (9)
eiq in equation (6). The most general eigenfunction for a given energy E is a linear
combination c1 qu( x2+) qc -(u x), with arbitrary complex coefficients c1 and c2 .
26
From this lesson one can get an clear idea on linear harmonic oscillator in detail. We
can calculate the energy eigenvalues. The Kronig Penny model, is discussed in
detail which find application field of nuclear physics and material science.
1. What are raisnig and lowering operators? Explain why they are so called.
2. Give the matrix theory of harmonic oscillator and obtain the energy eigenvalue
spectrum.
3. Explain Kronig Penny model.
4. Write down the matrices for the ladder operators a,a+ of the harmonic oscillator in
the representation which diagonalises the Hamiltonian.
3.9. References
1. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan. Tata
McGraw Hill Publishers.
2. Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
3. Quantum Mechanics V. Devanathan Narosa Publishing house.
4. Quantum Mechanics S.L. Gupta, V. Kumar, H.V. Sharma, R.C. Sharma. Jai
Prakash Nath & Co Ltd.
5. Quantum Mechanics Satya Prakash Sultan Chand Publishers .
6. Angular Momentum Techniques in Quantum Mechanics V. Devanathan
Kluwer Academic Publishers.
7. Quantum Mechanics John L. Powell and Bernd Crasemann Narosa
Publishing house.
.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
27
LESSON 4
STATIONARY PERTURBATION THEORY
4.1 Aim and Objectives
In this lesson, we are going to give an introduction to various approximation methods
and going to study Stationary perturbation theory and its applications in detail
4.2. INTRODUCTION
In quantum mechanics, there are many physical systems for which, the
equation of motion cannot be solved exactly. So, one has to resort to approximation
methods. There are three most important approximation methods in quantum
mechanics. They are
28
The stationary perturbation theory is concerned with finding the changes in the
discrete energy levels and eigenfunctions of a system when a small disturbance is
applied. It is assumed that the Hamiltonian H is the Schrdinger equation can be
written as sum of two parts.
i.e. H = Ho + H ' (1)
Hy =y
W
This watermark does not appear in the registered version - http://www.clicktoconvert.com
29
The assumption that H ' is small suggests that, we expand the perturbed
eigenfunction & eigenvalues as power series in H ' . This is easily accomplished in-
terms of a parameter l, such that, the zero, first, second.. powers of l correspond
to zero, first, second .. orders of perturbations. We replace H ' by l H ' and
express Y and W as power series in l. We assume that, these two series are
continuous analytic function of l for l between zero and one. The different orders of
perturbation approximation are then given by the coefficients of corresponding
powers of l. In the final result, l is set equal to one.
The perturbed wave function & energy level are written as
y = y 0 l+ y1 +l 2 2y +3l3 y+ . . . . . . . (3a)
W= W
0 +
l W l 2W
1 + 2 +l3 W
3 + ........ (3b)
using (2) and (3) in equation (1) we get
2 2
( H o' +)l ( .+y
H0. . l
+. 1 . Wy
. + . 2 . l =. W
)0 y( + . 1l. W
.+ l
. 2 . +. . . . )
2
(4)
(y 0. +l. y
1. +. l 2. y+
. . . . )
The above equation is valid for a continuous range of l . We now, equate the
coefficients of equal powers of l on both sides to obtain a series of equations that
represent successively higher orders of perturbation.
l 0 Hyo =0 W y
0 0
( H o- )W
0 y0= 0 (5a)
l 1 H oy +1H 0 =y 0 W1 + y1W
'W 0
( H 0- ) W0 y
=( W
1 -' 1 y
H )0 (5b)
s i m i l al r2 l
y f oo- r (0yH W
2= ) 1-
( W 1yH ' 2+) Wy
0
(5c)
l3 ( o-H W
0y )3= ( 1-W y
2 H 2+
' )y
1 W 2+ W y
0
(5d)
........................................
.........................
........................................
.........................
The first of the above set of equations means that y 0 is any one of the unperturbed
eigenfunctions. Therefore, we can write,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
30
y 0 u m= & o W
= mE (6)
This state u m is discrete, since we are dealing with the perturbation of a bound state. It
is assumed to be non degenerate as well, although, others of the unperturbed
eigenfunctions may be degenerate or continuous.
Expressing y1 =n a n
( 1 )
u n y &0 = mu
( H
o -W0 )an= (u1 )(n w- 1H ' m) u
n
( 1 ) ( 1
n an H 0 u- n a n )n u
W 0 = n u 1 -Hm '
W m u
n a(n E 1 n u-
) (
n a n
1
E n) u m = n u 1 -H m '
W m u
( 1 )
n a [ En -E m] u=n W1 u- mH 'm u
n
n a n( 1 [m)- E
n dk E n + ]1 dkW m =Hk m'
31
=n s ( iny2 n 1) g= ( na 1
E x p r ye2 s un ) , ny
a u0 =&m u
n
weget
n a (n H 20 u-n) a n( Wn2 u) = na(
0 n
1
W n
)
u 1 -(a n 1Hn ) ' n un+ W
2 m u
n a (n E 2n u-n) a n( En2 um) = an( 1 ) (
n n u 1 -a
W n
1
Hn) ' n un+ W
2 u
m
( 2 ) ( 1 ) ( 1 )
n a [ E -Em ] n -u 2 W m=u n a n -W 1 un a
n n n H n ' n u
Rearranging we get
n a( n [ E- En ) +]n u( 2= Wm1
m
2
un ) a n -H
( 1
n '
nu n a) 1 W n u
n a ( n[ E
-m 2 E n ])tku un * d+2 W*k u mtu d
= n a ( nu1 H
k
)
' n u*t n d
(
a 1 -Hm ) ' k u* tnu d
m
n
(
n a n [-m 2 nE ) dk E n( ]+1 k W)m d
2 na
'
k
(
=
H n
1
a
n -m)H m d' k n
n n
n m
where prime over the summation indicates that, we have omitted the term n = m.
Substituting the value of a (n 1 )from first order perturbation equation we have,
' | Hn m'2 |
W2 = n S
( mE- nE )
where we used the relation H m ' n= ( H n m' )*. The above equation gives
the second order perturbation to the energy eigenvalue.
For k m, equation (10) gives
This watermark does not appear in the registered version - http://www.clicktoconvert.com
32
a (k [ 2
m E -)k E (n =] 1n a )kSH(n '-1k a )
m H m '
u s i n g t h e v a l u e o(n 1f )a w e g e t
H ' H ' H ' H '
a (k 2 =) n' S k n n- m k m m m2
( mE - Ek ) m(- E n E ) ( mE- kE )
< k | H> ' | n <n | H '> | m k< | H ' >| m m<| H ' >
| m
a (k 2= )
n
'
S - 2
(12)
( mE - Ek ) m(- E n E ) ( mE- kE )
|< k | H >' | m
|<
' m | H>' | m | <k | H >
' | n | <
n | H ' >| m
y um = '
u+ k 1 k- + n (13 b)
E m - E n( E m-E )k ( m-E E k ) -(m E n E )
f t h e c o n d iyt0 i oy
tothesecondorderinH'.Thisoccuredbecauseo s n (=
, ) 0>, s 0
-e r12 -e
r1 r2
+Ze
NUCLEUS
Consider the normal state of helium atom which contains a nucleus of charge
Ze and two electrons each having a charge e. The potential energy of such a
system is given by
Z 2 e Z2 e 2 e
V= - - + (1)
r1 2 r 1 r2
in which r1 and r2 are the distances of electrons 1 and 2, respectively from nucleus
This watermark does not appear in the registered version - http://www.clicktoconvert.com
33
The first order perturbed energy E(1) is the average value of the perturbation
( )1
e2
function H = over the unperturbed state of the system. Hence
r1 2
e 2 ( 0 )2
E (1) = y n (H0 ) *yn( 1= ) ( 0 )
dt y d1 0 0 , 1t0 0 (6)
r1 2
we know that
Z3
u1(s0) 1= y00 ( )0= e 3
-r / 2
(7)
p a0
in which = 2Zr / a0 and a0 = h2 / (4 2 me2 )
This watermark does not appear in the registered version - http://www.clicktoconvert.com
34
Z 3 ( - r1/ 2 ) ( - r/ 2 )
y 1( 00)0 , 1 0=0 e e . (8)
p a03
The volume element is
2 2
dt = r1 dr1 sin q 1 dq 1 df1 r2 sin q 2 dq 2 df 2 (9)
So that the integral for E(1) becomes
Z 2e p
2p p 2 p e( - r1 - r)2
E (1) = d r12 r1
25 p 2a0 0 0 0
1
0r 0 0
2 (10)
2
s q1id q1n dj-
1 s rd i r
2 ndq2
2 q
2 d 2j
E ( =1 (
) 2
-
-2Z5/4Z ) H E. (12)
Thus we can calculate the energy of He, Li+, B e ++, etc. with Z = 2,3,4 etc.
respectively.
The ground state of hydrogen atom is non degenerate, but the n = 2 energy
levels of hydrogen atom is four fold degenerate. On the application of an external
electric field, the degeneracy is partly removed. The splitting of atomic energy levels
due to the application of external electric field is known as Stark effect and it has been
discovered as early as 1913. It is found that the splitting of energy levels of the
hydrogen atom for weak electric field is proportional to the field strength, and hence it
is known as Linear Stark Effect or First order Stark Effect.
The splitting of energy level for all other atom is proportional to second power
of the field strength and that is known as Quadratic Stark Effect.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
35
ground state y 1 0 0is non degenerate, but, the first excited state y 2l m
l
with n = 2 is
four fold degenerate. We shall write down explicitly the degenerate eigenfunctions,
u1 =2yR0 =0 r2q 0( Y0) 0 f( , )
u2 =2yR1 =0 r2 q 1( Y1) 0 f( , )
(1)
u3 =2yR1 =1 r2 q 1( Y1) 1 f( , )
u4 =2yR1 - =1 r2q 1(- Y1) 1f( , )
where, Yl m
l
(q , f) are the spherical harmonics and R21(r) and R20 (r) are the
radial part of the wave functions. The radial functions are given by ,
1 r - r / 2a0
R2 (0 = ) r -1 e (2)
2a03 2a0
1 r - r / 2a0
R2 (1 =
) r e (3)
2 6a0 a03
2
a0 = h 2 / m e (4)
The Hamiltonian of the hydrogen atom is taken as the unperturbed
Hamiltonian H0 and the applied electric field as the perturbation H'. If the electric
field E is along the z axis, the perturbation is given by
H' = - eEz
= - e- Er cos
4p
H 'e= Er 1 0 q jY(,) (5)
3
where, - e- is a positive quantity.
Since, H' is of odd parity, the matrix elements between the states of same
parity (same l value) vanish. The same result can be arrived at by considering the
angular integration involving the spherical harmonics. The only surviving matrix
elements are,
y2 0 H0y' 2 1 0
y2 1 H0y' 2 0 0
It can be easily seen that this two matrix elements are equal. Let us
evaluate one of them
This watermark does not appear in the registered version - http://www.clicktoconvert.com
36
y2 0 H0'y 2 1H =0*y2 d0 0 y
' 2r W1 0 d
4p
= e E 2 (R 0 ) r 2 ( R13) r1( Wr,0q d)0 f (r0 q, Y1 f) Y0 (q Y, f )d
3
1
where, Y00 (q , f =
)
4p
Y Y W
d =1
1 0 1 0
1
and hence the angular integration yields . The radial integral can be performed
4p
by substituting the radial function from equations (2) and (3)
-r
3 1 r 4 a0
R ( r) R( ) rr
0
2 0 2 1
0
=d r -1r e
4 3 2a0
d r
r
Letting x = the integral becomes,
a0
a0 x 4 - x
1 - x e d x
4 30 2
=
a0 5 !
= 4 -!
4 3 2
9a
= - 0 (7)
4 3
4p 91a
y 2 0 H0y' 2 1 0 e= . E - . 0
3 3 4p
4p
= e E
3
y2 H
0 y' 0 e 2 =
3 1E- 0 a0 (8)
H 2' because
1 of the definition equation (1) and denoting the quantity 3|e|E a0 by P1 we
have,
H1' =H' 2= P-2 =
e 1-
3E 0 a (9)
Substituting the values of H1' 2and H 2' 1in the secular equation,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
37
-W1 - P 0 0
- P -W 1 0 0
=0
0 0 -W1 0
0 0 0 -W1
If W1 = P, then C1 = - C2 and C3 = C4 = 0
If W1 = -P, then C1 = C2 and C3 = C4 = 0
If W1 = 0, then C1 = C2 = 0 and C3 and C4 are indeterminate
Then it follows that,
1
U1 = u1 ( - 2u )
2
1
U 2 = u1 ( + 2u ) (11)
2
U 3 =C3 u3+ C 4 4u
U 4 =C3 ' u3+ C
'
4 4u
give respectively the first order correction to the energy and the corresponding
zeroth order eigenfunctions. It is seem that out of the four degenerate states only the
two states y 2 0a 0y
n 2 d1 0 are split but the other two states y 2 1 and
1 y 2 -1 continue
1 to be
degenerate. This is diagrammatically shown in the following figure.
U1 W0 +3_e|Ea0
W 0 u1 u2 u3 u4
U3 , U4 W0
U2
W0 -3_e|Ea0
38
4.7. References
1. Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
2. Quantum Mechanics A.K. Ghatak and S. Lokanathan McMillan Publishers.
3. Quantum Mechanics Satya Prakash Sultan Chand Publishers
4. Quantum Mechanics V. Devanathan Narosa Publishing house.
5. Quantum Mechanics S.L. Gupta, V. Kumar, H.V. Sharma, R.C. Sharma.
Jai Prakash Nath & Co Ltd.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
39
LESSON 5
VARIATIONAL METHOD AND WKB
APPROXIMATION
is an upper limit to the lowest energy state of the system. In the above
equation, H is the complete Hamiltonian operator for the system and the function f is
any normalised function of the co-ordinates of the system. The choice of the function
f is quite arbitrary but the more wisely H is chosen, the more closely W will
approach the energy Eo . If we use the true wave function y o of the lowest state instead
of f , then W will be equal to E o .
It f is not equal to y o , we may expand f is terms of the complete set of
normalised orthogonal functions y o , y
, 1. . y2. . y
.n .
Substitutinginequation(1)weget
* *
W = a ya H y
n m n dm z
n m
* *
=
ay n a
m y n dm
E
m z
n m
*
=
a n dam mE n m
n m
= a *n an En
n
2
W = a n En (3)
n
Subtractingthegroundstateenergyfromthebo
thsides
This watermark does not appear in the registered version - http://www.clicktoconvert.com
40
2 2
W- E
0 =a n E-
n E Q a n =1
0
n n
2
= a n (- n E 0E ) (4)
n
2
As a n is positive and E n E0 , the R.H.S is positive or Zero. Thus the have proved
that W is an upper bound to the lowest value of energy E 0 .
QW E0 (5)
Note :-
If the function f is not normalised, then W is given by the expression
*
W=
f Hf dt
*
f fd t
The variation of f is performed by first choosing a suitable functional form for f .
Depending on a number of parameters l1 , . .l2. . . .ln . . . . .
. . ) f1* ( l, 2. .l. .n l
W (l1, . l. .2 .l. . n= . . . 1y) H l(2 , .l.2 . l. . )td
As the parameters are varied, the value of W also varies. Minimization of W is then
accomplished by finding the values of l1 , .l2. . ln. . for which
W W W
= . .=. . . . . . . .=. . . . . . . 0
1 l 2 l n l
41
1 -a r
2y = 2 2 r e
r r r
2
1 2 - ar
= 2 2r + r e
r r r 2
2
1 - ar 2
= 2 r e + r ( )e - ar ( )
r2 r r 2
1
=
r 2
[
2r (- a )e - ar + r 2 a 2 e - ar ]
2a
= e - a ra-2 (4)
r
According to variational method, the energy E is given by
E=
y * Hy dt
y * yd t
-a rh 2 -a 2 a 2r 2 - ea 2 r
0
e - a8p 2 m e- e - r
r r
d 4p r
=
e -2 a r4r p2 d r
0
where 4pr 2 dr is the volume of a spherical shell at a distance r and r+dr from the
nucleus. The limit of integration varies from 0 to as r goes form zero to infinity.
h2 2
2a - a
h2 r 2 2-a2a r 2 - a 2 r
- r e +d r r e - d r e r e d r
0 8p 2 m 08p2 m 0
=
(5)
2 -2a r
r e d r 0
n!
Applying the relation x n e - ax dx =
0 a n +1
The values of the above integrals can be obtained as
h2a2 2 h2a 1 2 1
- + -e 2
8p 2 m 8a 3 4p 2 m 4a 2 4a
=
2
3
8a
h2a2
= 2
- e2a (6)
8p m
Now we shall choose a in such manner as to give a minimum energy. Thus
E 2h 2
= -a e2= 0
a 8pm2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
42
4p 2 me 2
Or a= (7)
h2
Substituting the value of a in equation (6), we have
2
h 2 4p 2 me 2 2 2
2 4p me
E min = -e
8p 2 m h 2 h
2
2p 2 m 4 e
Em i n= - 2 (8)
h
This is same value as we have obtained by other methods.
43
quantum theory results and the higher terms, to the characteristics of the new
mechanics.
One dimensional Schrdinger equation is written as
-h2 2 y
+ ( -V y
E )= 0 (1)
2 mx 2
The WKB approximation takes advantage of the fact that the wavelength is changing
slowly, by assuming that the wave function is not changed much from the form it
would take of V were constant, namely
( i 2p x )
y e wh =h e r e=2 p 2 m- ( E V )
This suggests that, when E > V, it will be convenient to write, the wave function in
the form
( i S )
y =e h (2)
where S is a function of x (S may be complex).
It V is nearly constant, we may expect that s is roughly equal to p 2 x. Now, S can be
expanded as
h2
+h(1 x ) S+ (2 x ) . +. . . . . . . . . . .
S =S (o x ) S (3)
2
-1 ( S i
i h S / ) 2 h( i S / )S2
= e (+ ) e
h2 x h 2
x
equation(1)becomes
1 S 2 S 2 ( i Sh / )
- ih ( 2 + V E- ) =e 0
2 m x x
2
1 S S 2
- ih (+ V - E )= 0
2 m d x x 2
2
1 S i hS 2
+ (- V - E )=2 0 (4)
2 m d x m x
SubstitutingthevalueofSfromequation(3),( h
uptosecondorderin)inequation(4)weget,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
44
2
1 So S h 2S
+ h 1+ 2
2 m x x 2 x
ih 2 S h 2S 22 S
+( - V - 2oE h+ 2 1 )+ 2 2 =
0
2 x m x 2 x
2 2
1 So S S S 2SoS
+ h 2 1 + o2 h h12 + x
2 m
x
x
x
x x
i 2So 2S1
2
+( V E
- ) -h +h 2 =0
2 m x 2 x
2
1 So S0 i h S 2
+ (- V E +) -0 1
2 m
x m x x 2 x 2
2
(5)
h2 So S2 S1 2S
2
+ + - i = 0
2 m x x x x 2
o 2
1 So Si1 S
- = 0
m x x 2 x 2
S S Si 2
( o o r1 - o ) 2= 0 (7)
x x 2 x
1 So S2 S1 2 1S 2
+ - i = 0 ( o r )
2 m x x x x2
So S2 1S 2 S1
- -i 2 =0 (8)
x x x x
Fromequation(6)wehave,
So 1 2
= [ 2m (-E v] )
x
x
S0 [= 2 m ( E- V ) 1 ] 2 d x (9)
x0
i S
S1 = l o g o (10)
2 x
This watermark does not appear in the registered version - http://www.clicktoconvert.com
45
-1 / 2
-1 S S
i 1S= l o = go l oo g
2 x x
1 1
o r i 1eS = = 1 4
S o / x [ 2m (E- V)]
similarlyfromequation(8)wecanshowthat
1 m ( v / x ) 1 m 2( v /2 x )
(11)
2 [ 2 m -( E2 v3 ) ] 2[ -2 m5 ( E v2 ) ]
S2 = -d x
NeglectingS,thesolnofequation(1)isgivenb
2 y
i i 0S
( oS+ h S1 )
i 1S
y e h= e =h e
Asshaspositiveaswellasnegativevalues,
0
y A =( i0eSh / ) +
i 1
e S
B ( -ii 1S oh / S)
e e
i x
A +
= -e ( [ 2 m ( E V )1 ]2 d x )
h
[( 2 m ) -( E V] 1 )4 x0
i x
B -
+ -e (h [ 2 m ( E V )1 ]2 d x )
[( 2 m ) -( E V] )
1 4
x0
x x
A i ( p2 Bd x i ) 2( p- d x )
= e h x0
+ eh x0
(12)
p2 p2
where A & B are two arbitrary constants and P2 =2 m (- E V . )The positive exponential
represents a wave moving in the positive X direction while the negative exponential, a
wave moving is the negative X direction. When V is a constant, the exponential
reduces respectively to plane waves,
i i -
( 2p x ) 2( p x )
eh h
e
When V > E, the W.K.B approx possesses real exponential solution. The solution is
x x
i i -
A h x0 1 B hx0 1
p d x p d x
y =e + e
p1 p2 (13)
w h e r1 e= p 2 m- ( V E )
By reading this lesson one can get an understanding on the two approximation
methods namely variational method and WKB approximation and its applications.
Variational method is applicable to study the ground state of system. WKB method
find its application in alpha decay paradox.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
46
5.7. References
1. Quantum Mechanics S.L. Gupta, V. Kumar, H.V. Sharma, R.C.
Sharma. Jai Prakash Nath & Co Ltd.
2. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan.
Tata McGraw Hill Publishers.
3. Angular Momentum Techniques in Quantum Mechanics V. Devanathan
Kluwer Academic Publishers.
4. Quantum Mechanics John L. Powell and Bernd Crasemann Narosa
Publishing house.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
47
LESSON 6
TIME DEPENDENT PERTURBATION THEORY
AND ITS APPLICATIONS
6.1. Aims and Objectives
We are going to discuss
Time dependent perturbation theory and its applications
Fermi Golden rule
Harmonic perturbation
Constant perturbation
Selections rules
y
ih =
H y (1)
t
In many physical problems, the time dependence of the Hamiltonian arises due
to an external agency. For example, an atom irradiated by light, by applying electric
field to atomic system etc. In such cases, Hamiltonian is written as,
H =H 0 V
+ ( t ) (2)
where, H 0 is the sum of Hamiltonian of the two systems in isolation and V(t) is
their interaction.
At a particular instant of time t = 0, the system may be regarded as being
approximately in an eigenstate of H 0 as i. We have to calculate the probability of
finding the system at a later time t in an eigenstate f. If y n denotes the complete set
of eigen functions, then, we have
This watermark does not appear in the registered version - http://www.clicktoconvert.com
48
H 0 y n E=n yn
= h wn yn
The wave function corresponding to the actual state can be expanded in terms
of y n
y n ( t )= tn ) ( r )- i ewn t
nC ( y (3)
n
E - En
where, w s - n w = s is the Bohr angular frequency between the states s and n. If the
h
perturbation V(t) is absent, all the co-efficients C n will be constant in time. On the
other hand, the perturbation V(t) makes them vary with time and for this reason, this
method is called the method of variation of constants.
Suppose the perturbation begins to act at time t. Initially at t = 0, the system is
in the unperturbed state say, i before the perturbation is turned on. In such a case,
Cn ( 0=) 0
f o r n i
Cn ( 0=) 1
f o r= n i
1
Cf =Vf ( i t ) +ei wf ti
t ih
Integrating the above equation between the limits 0 and t, we get,
t
C f t 1 iw t
0 t d 0=tihV (f ti ) ef i d t
t
1
Cf ( ott =) fV (i t ) iwef ti d t
ih 0
t
1
Cf ( t-)f C (=f O ) iV ( t i)wf eti d t
ih 0
Q Cf ( 0= ) 0
t
1
Cf ( t =f) V i( t ) eiwfdti t , f o r f i
ih 0
This gives the probability amplitude for states of later time other than the
initial state.
49
Pf ( t =)f | C ( t2 ) |
| fV (i 2 o f 2) wi| t2
= s i n
/i 2
f
h2 wf
If the level spacing is very close, then, the summation can be replaced by integration.
= f
r( Ef ) df E
where r ( fE )is the density of the final states around E f (i.e) the number of energy
levels per unit energy interval.
| V (2 0 ) | tw / 2
Therefore, Pf ( = t ) f r2( f Ei ) f s 2 ifi n d E
h w f /i 2
| f V 2( i 0 f) | 2 i tw / 2
= r ( fE ) s i f n d E
h2 w f /i 2
A s s u m i n g e ( fE ) d o e s n o t vfa r y a r o u n d E , w e h a v e ,
| V 2( 0 ) | t w / 2
Pf ( = t f )2 r( f i E )f s 2 iif n d E
h - wf /i 2
| V 2( i 0 f) | 2 it w / 2
Pf (= tf f) r( E f )w s ii = hn dQ [w hE ]
h2 - wf /i 2
| V 2( 0 ) | t w / 2
Pf (= t f 2)f (i rE hff) s2 i w
i in d (1)
h - w f /i 2
Since the integral is practically zero outside w f ,i the limits are chosen from - to + ,
consider, the integral,
+
2 w f ti / 2
s
-
i w n/ 2 dw f
f i
i
p u =
t xw
f ti / 2
dw f it
d =x
2
2 w+ f ti / 2 s i2 n x 2
s w w f /i - 2
i n = df i . d x
-
( x /t 2t )
+
2 2s i n2 x
t - x 2
= t d x
+
s i2 n x
= 2 x 2d x
t
-
+
s i2 n x
we can show that d =x p
-
x2
+
2 w t / 2
s i f ni d w f 2 = t p
- w f /i 2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
50
substitutingtheintegralvalueonequation(1
)weget,
2
V ( 0 )
Pf ( =t f) i f (r E )p 2 t
h
2p t 2
= Vf r ( 0i ) f( E )
h
-
ThisiscalledtheFermiGoldenrule.
t
1 e iwf( i t )
= Vf (i 0 )
ih i wf i 0
Vf (i 0 )iwf ti
Cf (= t -) e- 1
hwf i
Thus, the probability of finding the system, in the state f at time t is given by,
2
Pf ( t=)f C ( t )
V 2f ( i 0 )
Pf =( t -)
h 2 w 2f i
( iwf
) ( e-
e t i - 1 if w it
1 )
w h e
( iwf
) ( )
r - tei - ,i f ew -ti 1 wif =e it -
1 - 1f i w-ei t e+ 1
= 2 - -ei f w+t i e- if wi t
= 2 2-c o ws f (i t )
= 2 [ -1 c owsf (i t ) ]
w t
= 2 2 s i2 n f i
2
w t
= 4 s i2 n f i
2
2
V ( 0 ) w t
\ Pf (= t )f2 i 42 s i 2n f i
hwf i 2
2 2
Vf (i 0 ) 2 wf ti / 2
= s i n
h2 wf ti / 2
51
Thereforethetransitionprobability,
2
2 V 2 ei wf ( - )1i -wt f wi e(+i -)w t 1
Pf ( t=f) c ( t ) =02 +
h (wf - )i w f w( i + )w
We find that the terms inside the box, while squaring will not contribute
anything because when w f i approximately equal to w only one term (IInd term) will
contribute while the other (Ist term) vanishes. Similarly, when w f i approximately
equals - w , the contribution by the two terms will exchange. Thus, the above
discussion treating the terms independently is valid.
We know that the probability amplitude for transition from initial state i to
the final state f will be,
t
1 iwf( i t )
Cf ( =
tf ) V i ( t ) e d t
ih 0
Vo t i
= w e -(f w
) t
ih o
i
t
V ei w(f i-) w t
= o
ih i wf ( i- w) o
Vo
= [ i ( wef -)i wt1
- ]
ih w
i f i-( w)
V ei ( wf -)i w1t-
= -o
h wf i- w
52
In this case we find that E f is lower than Ei by hw and this describes the emission of
energy, Here, the energy is given by the system to the external agency.
f i n a l f i n a
i n i t i a l i n i t i
Vf ( ti )iwfef =V
ti wif it
i ( o ) e
whereV(0)isindependentoftime
f i
\e q u a t i o n ( 4 ) b e c o m e s
This watermark does not appear in the registered version - http://www.clicktoconvert.com
53
t t
C f 1 iw t
0 t i h0V f( i o ) ef i d t
d =
t
t
1
Cf ( t=f ) Vi ( o )iwfeti d t
ih 0
This equation shows that Vf (i tis) non- zero only when Vf (i o is) non-zero, Thus,
transitions in the first order, can take place between pair state(f,i) for which Vf (i o ) .0
Transition between any other state are forbidden in the first order. This is the first
order selection rule. Even if Vf (i 0= ) then,
0 transition between the state i and f may still
be possible in the higher order via an intermediate ( l ) state. If both Vfl ( 0 and
)
Vli ( 0 ) ,0then the intermediate state l will connect the initial state i and final state f.
However, probability for transitions in the higher order will be smaller and will
proceed at a slower rate.
If we plot the probability Pf ( t as) a function of the energy difference w f i
between the initial and final state of fixed t, the graph will be of the form as shown in
the figure below,
P(t)
f
p t 3 / 5p/ t
p t 5 / p t 3 /
w t / 2
The behaviour of sin s i n f i is depicted in the figure.
w f i/ 2
54
6.9. Problems
Calculate the second order energy correction for the ground state of the charge
between linear harmonic oscillator subject to a constant electric field. Compare the
result with the exact solution
This watermark does not appear in the registered version - http://www.clicktoconvert.com
55
Solution
The schrodinegers equation for the perturbed system is
-h 2 d 2y 1 2
+ k y+ x a x=y E y
2m d 2 x2
where is the product of the charge and the magnitude of the electric field in the x-
direction. Writing X = x + / k, the above Equation becomes
-h 2 d y 1 2 2 a1 2
2
+ k =yX E+
2m d2 X 2 k
1 1 a2
Thus En = n + hw- , which is the exact result.
2 2 k
First order energy correction is zero.
The second order energy correction is given by
2 ( 0 ) ( 0 2)
'a n x m
En( 2 =) 2 m
(E( 0 ) ( 0 )
n -E
m )
( ( 10 0h) )( 0 0) )( a 0 )
n0 x ) =m( n a -( m n m
i 2mw
1 h
= - md n , +m dn ,+1 m 1
i 2m w
- m1
a2 a2
= En( 2 =) - = -
2mw 2 2 k
Do it yourself
1. A hydrogen atom the energy state with n = 2 is subject to an electric field E
parallel to the z axis. Calculate the energy correction.
2. Calculate the splitting of hydrogen energy levels due to spin-orbit interaction.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
56
6.10. References
57
LESSON 7
ANGULAR MOMENTUM ALGEBRA
7.2. Introduction
Knowledge in angular momentum techniques is essential for quantitative study
of problems in atomic physics, molecular physics, nuclear physics and solid state
physics. The application of angular momentum techniques to solve physical problems
has become so common. Angular momentum techniques are widely used in the study
of both non relativistic and relativistic problems in physics. In this chapter, we will
discuss the Quantum Mechanical definition of Angular m omentum first. Then we
will have a discussion on physical interpretation of Angular m omentum. It is then
followed by, discussion on ladder operators and coupling of two angular momenta.
i j k
u r
L= x y z
px py zp
r
L = (y pz z py ) i (z px x pz) j + (x py y px ) k
This watermark does not appear in the registered version - http://www.clicktoconvert.com
58
i.e. Lx = y pz z py
Ly = z px x pz
Lz = x py y px
[Lx , Ly ] = [y pz z py , z px x pz]
= [y pz , z px ] [y pz , x pz] [z py , z px ] + [z py , x pz]
= y px [pz, z] + py x[z, pz]
= i h y px + i h p yx
= i h [x py y px ]
= i h Lz
i.e. [Lx , Ly ] = i h Lz (2)
Similarly
[Ly , = i h Lx
Lz ]
[Lz , Lx ] = i h Ly
r r
i.e. ( L x L ) = L x Ly Ly L x
r r
(LxL) = i h Lz
( J x J ) = i J (4)
Now, [Jx , Jy ] = i h Jz
[Jy , Jz] = i h Jx (5)
[Jz , Jx ] = i h Jy
59
[J2 , Jy ] = 0 (6)
[J2 , Jz] = 0
Equations (5) & (6) gives simple physical interpretation. It is possible to find
the simultaneous eigenvalues of J 2 and one of the components, say Jz alone, but, it is
impossible to find precisely the eigenvalues of Jx and Jy at the same time.
2
Representing the operators by matrices. One can say that, J and Jz can be
diagonolised in the same representation, but not the other components Jx and Jy .
Physically this means that, one can find the magnitude of angular momentum vector
and its projection on one of the axes. The projection on the other two axes cannot be
determined. This is illustrated in the figure(Fig 7.1) , in which the angular momentum
vector is depicted to be anywhere on the cone. If y jm is the eigenfunction of the
r
operators J 2 and Jz, then
r2
J y jm = hj y jm
& Jz y jm = m y jm
Here j & m are Quantum Numbers used to define eigenfunctions and the
corresponding eigenvalues of the operators are hj & m.
Z axis
Jz
r r
J J
60
Problem 7.1
Show that [J2 , Jx ] = 0
Solution
r r
We know that J2 = J x J
= Jx 2 + Jy 2 + Jz2
\ [J2 , Jx ] = [Jx 2 + Jy 2 + Jz2 , Jz]
= [Jx 2 + Jx ] + [Jy 2 + Jx ] + [Jz2 + Jx ]
= Jx [Jx , Jx ] + [Jx , Jx ] Jx + Jy [Jy , Jx ] + [Jy , Jx ]Jy
+ Jz [Jz , Jx ] + [Jz , Jx ] Jz
= i h Jy Jz i h Jz Jy + i h Jz Jy + i h Jy Jz
i.e. [J2 , Jx ] = 0
Similarly [J2 , Jy ] = 0
& [J2 , Jz] = 0
61
= 0
Similarly [J2 , J ] = 0
Also [J+ , J2 ] = 0
and [J , J2 ] = 0
Problem 7.3
Find the value of [Jz , J+] and [Jz , J ]
Solution
Now, [Jz, J+] = [Jz , Jx + i Jy ]
= [Jz , Jx ] + i [Jz , Jy ]
= i h Jy + i ( i h Jx )
= i h Jy + h Jxs
= h (Jx + i Jy )
= h J+
And, [Jz, J ] = [Jz , Jx i Jy ]
= [Jz , Jx ] i [Jz , Jy ]
= i h Jy i ( i h Jx )
= i h Jy h Jx
= h (Jx iJy )
= h J
Problem 7.4
Find the values of Jz J+ and Jz J
Solution
We Know that, [Jz , J+] = J+ in units of h
i.e. Jz J+ J+ Jz = J+
or Jz J+ = J + Jz + J +
i.e. Jz J+ = J+ (Jz + 1)
Similarly [Jz , J ] = J in units of h
i.e. Jz J J Jz = J
This watermark does not appear in the registered version - http://www.clicktoconvert.com
62
Jz J = J Jz J
i.e. Jz J = J (Jz 1)
63
Similarly Jz j = Jz J y jm
= J Jz y jm
= J (Jz 1) yjm
= J (m1) y jm
= (m1) J y jm
i.e. Jz j = (m1) j
Thus we find that j+ and j are eigenfunctions of J2 operator, with the same
eigenvalue hj. However, even though we find that, j+ and j are also eigenfunctions
of Jz operator, the eigenvalues are either stepped up or stepped down by unity. It is
precisely for this reason, J+ and J operators are called as raising and lowering
operators, respectively.
Now let us find the value of (J2 x + J2 y ) y jm
(Jx 2 + Jy2 ) y jm = (J2 Jx 2 ) y jm
= (hj m2 ) y jm
Now, let us denote the lowest value of m by m1 , and the highest value of m by
m2 .
Obviously J+ jm = 0 (A)
2
and J jm = 0 (B)
1
Operating J on the left of equation (A) and J+ on the left of equation (B), we have
J J+ jm 2 = 0
[J 2
x ]
+ J 2y - i (J y J x - J x J y ) jm 2 = 0
(J 2
x + J 2y - J )
z jm 2
= 0
(J 2
- J 2z - J )
z jm 2
= 0
This watermark does not appear in the registered version - http://www.clicktoconvert.com
64
[J 2
]
- J z (J z + 1) jm 2 = 0
[ j - m 2 (m 2 + 1)] jm 2
= 0 (C)
Similarly J+ J jm1 = 0
[ j - m 1 ( m 1 + 1) jm1 ] = 0 (D)
Since R.H.S of (E) and (F) are equal to zero, we can write,
j - m 2 ( m 2 - 1) = j - m 1 ( m 1 - 1)
i.e. m 2 (m 2 - 1) = m1 (m1 - 1)
m 22 - m12 + m 2 - m1 = 0
(m1 + m 2 ) (m 2 - m1 + 1) = 0 (G)
Since (m 2 - m1 ) is always > 0, (i.e positive), the second term cannot be zero, we have
(m1 + m 2 ) = 0
m1 = - m2
65
= y * jm+1 |+ |2 y jm+1 d
= |+ |2 y* jm+1 y jm+1 d
= |+ |2 (3)
Using equations (1), (2) and (3) we can write,
|+ |2 = j(j+1) m(m+1)
or + = +
j(j1) -+ m(m1)
= j2 + j - m 2 - m
= (j - m) (j + m + 1)
In a similar fashion, we can show that,
= (j + m) (j - m + 1)
(j + m) (j - m + 1) respectively.
The above matrix elements are sufficient to construct all the angular
momentum matrices.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
66
Note
1. 1
Jx = (J + + J - )
2
2. 1
Jy = (J + - J - )
2
3. Since all the matrix elements defined above connect states with the same j but
different m values, it is usual to construct angular momentum matrices for a
given j value.
4. It is customary to label the rows by m' values (i.e. m' = j..j) and the
columns by m values (i.e. m = j..j)
5. For a given j, the dimension of the angular momentum matrices are given by
(2j+1) X (2j+1)
Problem 7.5
Find the angular momentum matrices of J2 , J z , J + , J , Jx and Jy
when j = .
Solution
Dimension of the matrix
The dimension of the matrix is (2j+1) X (2j+1) i.e. (2 X +1) X
(2 X +1) = 2 X 2. Then, the possible values of ma are and .
The angular momentum matrix for J2 is given by.
m'\m 1
2
-1
2
1
2 3
4 0
-1 0 3
2 4
3 1 0
=
4 0 1
The angular momentum matrix for Jz
m'\m 1
2
-1
2
1
2 12 0
-1 0 -1
2 2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
67
1 1 0 1
= = sz
2 0 - 1 2
For J+
m'\m 1
2
-1
2
1
2 0 1
-1 0 0
2
For J
m'\m 1
2
-1
2
1
2 0 0
-1 1 0
2
1 0 1 0 0
= +
2 0 0 1 0
1 0 1 1
= = sx
2 1 0 2
Similarly for Jy
1
Jy = (J+ + J )
2i
1 0 1 0 0
= -
2i 0 0 1 0
1 0 1
=
2i - 1 0
1 0 1
=
2 - 1 0
1 0 - i 1
= = sy
2 i 0 2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
68
Note
1 0 1 1
Jx = = sx
2 1 0 2
1 0 - i 1
Jy = = sy
2 i 0 2
1 1 0 1
Jz = = sz
2 0 - 1 2
are the Paulis spin matrices for j = .
Problem 7.6
Obtain the angular momentum matrices J2 , Jz , J+ , J , Jx and Jy for j=1.
Solution
Dimension of the matrix (2j + 1) x (2j + 1)
= (2 x 1 + 1) x (2 x 1 + 1)
= 3 x 3 matrix
j = 1 possible values of m are 1, 0, 1.
Angular momentum matrix for J2
m'\m 1 0 -1
1 2 0 0 1 0 0
0 0 2 0 = 2 0 1 0
-1 0 0 2 0 0 1
69
m'
m 1 0 -1
1 0 0 0 0 0 0
0 2 0 0
= 2 1 0 0
- 1 0 2 0 0 1 0
0 1 0
2
= 1 0 1
2
0 1 0
0 1 0
1
= 1 0 1
2
0 1 0
0 - 1 0
i
= 1 0 - 1
2
0 1 0
This watermark does not appear in the registered version - http://www.clicktoconvert.com
70
Problem 7.7
3 3
Dimension of the matrix = 2x + 1 x 2x + 1 = 4 x 4 matrix
2 2
3 1 -1 3
\ possible values of m are , , , .
2 2 2 2
15 3
2 0 0 0
4 0 0 0
15 0 1
0 0 0 0 0
4 2
J2 = Jz =
0 0 15 0 0 -1
0 0
4 2
15 3
0 0 0 0 0 0
4
2
0 3 0 0 0 0 0 0
0 0 4 0 3 0 0 0
J+ = J =
0 0 0 3 0 4 0 0
0 0 0 0 0 0 3 0
0 3 0 0 0 - 3i 0 0
1 3 0 4 0 1 - 3i 0 - 2i 0
Jx = Jy =
2 0 0 0 3 2 0 2i 0 - 3i
0 0 0 0 0 0 3i 0
LET US SUM UP
By going through this chapter, one can easily understand the algebra of
angular momentum. We can calculate the spectrum of eigenvalues for J2 and J z
operators. One can easily construct the angular momentum matrices for a given j by
studying this lesson.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
71
72
LESSON 8
COUPLING OF TWO ANGULAR MOMENTA OR
CLEBSCH GORDON (C,G) COEFFICIENT
r r r
When we couple two angular momenta J1 and J 2 , the final J will be between
|J1 J2 | in steps of unity. We will have two sets of mutually commuting operator
given by
Set I : J12 , J2 2 , 2 J z , J
The state vector corresponding to the first set j1 j 2 j m is called the coupled
j1 j2 j
j1 j 2 j m = m m2 m
j1 j 2 m1 m
2 (2)
m1 m2 1
j1 j2 j
The symbol is called the unitary transformation coefficient
m1 m2 m
73
Now, taking scalar product with j1 j2 j' m' on both the sides of equation (1)
j1 j 2j' m1 ' j2 j 1m 2m
j1 2j j
j1 j 2j' m1 ' j2 j 1m 2m = m m2 m
j1 j 2j' m1 ' j2 j 1m 2m
j m 1
j1 2j j
= m d d
m2 m j 'j m 'm
j m 1
Now, doing the summation over j and m on R.H.S, because of the presence of
Kronocker delta, of all the possible values of j and m, only those terms corresponding
to j = j' and m = m' will remain.
j1 j2 j
j1 j 2j m1 j2 j 1m 2m = m (3)
1 m2 m
Thus we find the C.G. coefficient is the scalar product of the two state vectors
in the coupled and the uncoupled representation.
In a similar manner, taking the scalar product with | j1 j2 m1' m '2 > on both the
sides of equation (2),
j1 j2 j
j1 j 2 m1 m2 j1 j2 j m = m (4)
1 m2 m
Taking hermitian adjoint (Congugate transpose) of equation (4)
j1 j2 j
j1 j 2j m1 j2 j 1m 2m = m (5)
1 m2 m
j1 j 2 m'1 m'2 j1 j2 m1 m
2 = d m m' dm ' m (2)
1 1 2 2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
74
j1 2j j
j1 j 2 j m = j1 j 2 m1 m
m1 m2 m 1
2m m
2
j j j' j1 j2 j '
m1 m2 1' 2
' m m m d m1 m' 1dm2 ' = d jj' dmm'
m m2' m 1 2
m2
m1' m2' 1
j1 j 2 j 1 ' j 2 j j
= d j 'j dm 'm (3)
m1 m 2 m 1 m 2m '1 'm 2 ' m m
In a similar way, starting from equation (2), the orthogonal property for the
C.G. coefficient can be written as
j1 2j j j1 j j
2
= d m m' d (4)
j m m '1 m' 2 m ' m1 m2 m 1 1 m 2 m '2
3. [ j] j1 j j2
= (-1) j1 -m1
[ j2 ] m1 - m - m 2
4.
( - 1 j2 + m)2
[ j] j j2 1j
= -m
[ j1 ] m2- m
1
where [ j ] = (2j+1)1/2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
75
1. Triangle Rule
D (j1 j2 j) must be satisfied. i.e. the final j must be within the permissible limit
of |j1 j2 | to j1 + j2 (in steps of unity).
2. m sum rule must be satisfied
m 1 + m2 = m
3. Projection rule must be satisfied i.e. m should be valid projection of j. i.e. m = j,
j +1 .(j1), j is steps of unity.
2
j1 j2 j
= 1 (5)
m1m 2 m1 m2 m
2
j1 j j
2
= 1 (6)
j, m m1 m2 m
1. When either j1 or j2 is zero, then the coupled and uncoupled representation, will
become equal. In such a case, the corresponding C.G. coefficient will be equal
to unity.
0 j j
i.e. 0 m m = 1
j 0 j
m 0 m = 1
76
We have studied the coupling two angular momenta in detail in this chapter.
Also we discussed the symmetry properties of C.G. coefficient.
8.9. References
1. Angular Momentum Techniques in Quantum Mechanics V. Devanathan
Kluwer Academic Publishers.
2. Quantum Mechanics V. Devanathan Narosa Publishing house.
3. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan. Tata
McGraw Hill Publishers.
4. A text book of Quantum Mechanics S.L. Kakani Kedarnath Publishers.
Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
77
LESSON 9
RELATIVISTIC QUANTUM MECHANICS
9.1. Aims and objectives:
In this lesson we will be studying the important aspects of relativitic quantum
mechanics such as Klein Gordon equation and Diracs relativistic wave equation.
Then we will have a discussion on the free pariticle solution for Diracs equation.
Also we are going to study the properties of Diracs matrices.
9.2. Introduction
The Quantum Mechanics developed so far, does not satisfy the requirements
of special theory of relativity as it is based on a non relativistic Hamiltonian. The
wave equation we used to discuss physical problems is of the first order in time and of
second order in space co-ordinates. This contradicts the space-time symmetry
requirement of relativity. The difficulties suggest two ways of obtaining relativistic
wave equations. i) to use a relativistic Hamiltonian for a free particle is place of non-
relativistic one ii) to try for a first order equation in both space and time co-ordinates.
Though Schrdinger himself suggested the first possibility its implications were
studied in detail by Klein and Gordon in 1926. Hence it is popularly known as Klein
Gordon (K.G.) equation. It does not say anything about the spin of the particle. Hence
it is used to describe spin zero particles such as in pions, kaons etc. Based on the
second idea, Dirac developed his relativistic wave equation. The concept of spin has
evolved automatically from Diracs theory which is applicable to spin systems.
78
c 2 p 2 + m2c 4 y = i h y
t
But it is very difficult to give a physical meaning for the square root of an
operator. Further if the square root is expanded in terms of binomial services we will
have even powers of on the L.H.S. In order to avoid the difficulties Klein-Gordon
proposed that the wave equation will be
H2 y = E2y
2
{c 2 2
p+ 2 4
}
y c ih y
m =
t
u 2r 2
{ 2
(
c - ih 2m + 4c h=
y ) } 2
- y2
t
u 2r 2
{ - h 2 c 2m + 2 ch =y4 } -2 y 2
t
Dividing throughout by - h 2 c2
2 m2 c2 1 2
- = y y
h2 2
c 2 t
2 2
2 1 m c
-2 2
=2 - y 0
c t h
m 2c 2
{2 - } y =0 (1)
h2
where the DAlemberts operator, which is relativistically invariant
1 2
2 = 2 -
c 2 t 2
Equation (1) is relativistically invariant, if y transform like a scalar.
79
d2 f
substituting y ( x, t ) = f (t ) e ik . x . This leads to 2
= [c 2 +
(k2 / m
2
)h2c ] f. Solving this
d t
we obtain, the plane wave solutions
i ( k m.w
x t ) ih (p x . -E ) t
y ( x , =)t e= e (2)
1
E = hcw =p ( 2 m2 + 2) c 4 2 (3)
In contrast to the nonrelativistic case where the coefficient ( E h ) of ( it) in
the exponent is the positive quantity ( p 2 / m2h ,) here we have solutions with -w as
well as with +w. The appearance of the negative energy solutions is typical of
relativistic wave equations.
Another difference from the nonrelativistic case is that y * y cannot be
interpreted as the probability density P(x,t). We expect P(x,t) to satisfy a continuity
3
equation so that P ( x, t ) d x is time independent. To obtain such an equation we
multiply equation (1) on the left by y*, its complex conjugate equation by y, and
subtract. The resulting equation can be written as
1 2 y
2 2
y* - (y * y - yy *) = 0
c t t
This is the continuity equation with
ih y y *
P ( x, t ) = 2
y* -y (4a)
2mc t t
ih
S ( x ,t= )- y( y*- y
* y ) (4b)
2m
A convenient choice of a common constant factor in P and S has been made
here. With this choice, S coincides exactly with the corresponding nonrelativistic
expression. However, P is quite different. It vanishes identically if y is real, and in the
This watermark does not appear in the registered version - http://www.clicktoconvert.com
80
y
case of complex wave functions, P can even be made negative by choosing
t
appropriately. Clearly, P cannot be a probability density. One could multiply P by a
charge e and then interpret it as a charge density. However, P can still have different
signs at different points, and this is hardly satisfactory in the description of a single
particle of given charge. It is known that equation (1) can be used to describe a system
of arbitrary numbers of particles and their antiparticles by treating y itself as an
operator function instead of a numerical valued function. At this level, the above
mentioned properties of P are no longer objectionable.
When the wave equation is of first order in time, it must be first order in space co-
ordinates too. Dirac, was probably influenced by Maxwells equation as they are first
order equations in both space & time co-ordinates. In 1928 Dirac proposed a
Hamiltonian that is linear in mass and energy given by
uurr
(
H = ca . +mpb c 2)
Dirac equation will be
u r r
Ey = { a( c.+ )pb }
y2 mc
u r r
{E -(ac .) bp 2
-y }
=
mc 0 (1)
81
82
Tr (b 2ax ) = - Tr(ax )
Tr(a x ) = - Tr(ax )
2 Tr(a x ) =0
Tr(a x ) = 0
In a similar manner we can prove that
Tr(a y ) = 0
Tr(a z) = 0
Tr(b) = 0
6. Since the traces of all these four matrices are zero, and their eigenvalues are
+1 and -1, there should be same number of +1 along the diagonals as the
number of -1 in b matrix. Therefore the dimension of b matrix must be even.
7. The ith element of the equation
axb + bax =0
i.e. (a )
k
x b kl + (b )
m
(a x )ml = 0
ik im
83
ie a x1 a x2 = 1 and a x2 ax1 = 1
We choose both, a x1 and ax2 to be sx
0 0 0 1
0 s x 0 0 1 0
ax = =
s x 0 0 1 0 0
1 0 0 0
84
It is to be mentioned that the above form for a and b is not unique. Any other
set can also be chosen to satisfy the above commutation relations.
However these two sets will be related by the singular transformation
a = s -1 a s
b = s -1 b s
where S is a non-singular free particle solution
t
r r u r r r
=
ih j u w(-i)
i ( k -. w
er )t
{ -ih a c( i )k b+ u }jmc
2 i ( k . r- wt )
e
hw uj = { c (a . hk ) + b mc }u 2
j
r r
Since hw = E h =
and k we
p have
urur
E uj =
{ c (a. p ) +b 2
mc} j u
u1 u1
u u
E 2
u 3
= {
c a x p x+ ay y p+ az + bp
z
2
} mc2
u3
u 4 u 4
This watermark does not appear in the registered version - http://www.clicktoconvert.com
85
u1 0 0 0 1 0 0 0 - i 0 0 1 0
u 1
0 0 0 0 0 i 0 0 0 0 - 1
E 2 = c
x + p
+ p + p
u 3 0 1 0 0 0 - i y 0 0 1 0
z
0 0
1
u 4 0 0 0 i 0 0 0 0 - 1 0 0
1 0 0 0 u1
0 1 0 u
0 2 2
mc
0 0 - 1 0 u3
0 0 0 - 1 u 4
Adding all the matrices
m 2 c c 0 p c( z p x) i -y p u 1
0 m (2 c c px +) i y p - cz pu 2
=
c p c ( px - iy p)- m c2 0 u 3
c( x +p y i )z p- c 0p - m cu2
4
Since the equation are homogeneous in uj the solution will exists only if the
determinant of the coefficient is zero.
E - m2 c 0- c p -z c p-
0 E - m 2 c- c + p c pz
=0
-c z p -c p- E+ m c2 0
-c + p cz p+ 0E m c2
By expanding the determinant we can show that
E2 = c2 p2 + m2 c4
2
E= cp+2 2 4
mc
This watermark does not appear in the registered version - http://www.clicktoconvert.com
86
Thus we find that both positive and negative energy values are possible for the
Dirac particle corresponding to E+ and E- we will have two solutions for both
2
Taking first E+ = + cp 2 + 2 4
, there are two linearly independent
mc
Case (i)
u1 = 0 u2 = 1
c -p
From equation (C) u3 =
E + m 2c
-c zp
From equation (D) u4 =
E + m 2c
0
1
\ The wave function will be
c -p/ E +(m 2c)
-c z p/ +E ( m 2c)
Case (ii)
u1 = 1 u2 = 0
c zp
From equation (C) u3 =
E + m 2c
c +p
From equation (D) u4 =
E + m 2c
1
0
\ The wave function will be
c z p/ E +(m 2c)
2
c +p/ E +(m c)
2
Similarly if we choose the negative root, E=-
- cp 2 + 2 4
, there are two
mc
Case (iii)
u3 = 0 u4 = 1
c -p
From equation (A) u1 =
E - m 2c
This watermark does not appear in the registered version - http://www.clicktoconvert.com
87
-c zp
From equation (B) u2 =
E - m 2c
c -p/ E -(m 2c)
-c z p/ -E ( m 2c)
\ The wave function will be
0
1
Case (iv)
u3 = 1 u4 = 0
c zp
From equation (A) u1 =
E - m 2c
c +p
From equation (B) u2 =
E - m 2c
c z p/ E -(m 2c)
c +p/ E -(m 2c)
\ The wave function will be
1
0
0
c + p c - pz 1
N2 = 01 = 1
E + m 2c +E 2m c c -P/ E+m 2 c
2
-c zP/ +E m c
2 c 2 + pz 2p cz 2 p
N 1+ + = 1
( E+ m2 c )2+ ( E2 m 2 c )
2 c 2 xp 2+ y 2p c 2 zp2
N +1 + = 1
( E + m2 c )2+ ( E2 m 2 c )
c 2 p2
N 2 1+ 2= 2 1
( +
E m c )
Therefore the normalized wave function will be
This watermark does not appear in the registered version - http://www.clicktoconvert.com
88
0
1 r r
1 i(k. r w- t)
y1 = e
2
c p 2 c -p/ E +(m 2c)
1+
( E + m ) 2c 2 -c z p/ +E ( m 2c)
In a similar manner for the other three cases also the normalization constant
can be shown to equal to
1
0
1
y2 = 2
2 2
c p cp z /( E + mc )
1+ 2
( E + mc 2 ) 2 cp + /( E + mc )
c z p/ E -(m 2c)
2
y4 =
1 c +p/ E -(m c)
c 2 p2 1
1+
( E - m ) 2c 2 0
E = c2p 2 + m 2c4
89
It was found that necessary condition for the solution of Dirac equation is
E= 2cp 2+ 2 4
mc
The meaning of positive energy state is clear but that of negative energy is not
clear. It was at one time, suggested by Schrdinger himself as having no meaning and
hence to be excluded. However there were two fundamental objections for excluding
the negative energy states,
i) Mathematical and another
ii) Physical.
The exclusion of negative energy states will give rises to an incomplete set of
solutions and this is mathematically not valid. On the otherhand the Dirac equation
starts with positive energy states and there is a possibility of induced transition in to
the negative energy states. Therefore if the negative energy states are excluded there
will be contradiction, from Physicsts point of view. This has led Schrdinger into
insurmountable difficulties.
In order to overcome the above difficulties, Dirac postulated that
1. The positive energy levels form a continuum from +mc2 to + where as the
negative energy states form another continuum from mc2 to -.
2. In between mc2 and +mc2 there are no available energy levels.
3. Dirac proposed that all the negative energy states are completely filled up
normally.
4. The electrons in the negative energy states will not be affected by electric and
magnetic fields.
90
itself, as a hole in the negative energy states. This hole is regarded as positions
and it behaves like an electron with the positive charge.
3. The fact that the negative energy states are completely filled up implies that
Paulis exclusion principle is obeyed. If the Paulis exclusion principle is not
valid, the negative energy states cannot be completely filled up. Since
otherwise many electrons rather than one electron will occupy level.
4. Feymann interpreted that negative energy states are electrons moving forward
in time (or) positions moving backward in time.
1. If you were a Dirac How would you have predicted the existence of positron?
Explain
2. The K.G. equation has nothing to say about the spin of the particle Illustrate.
3. Explain how Dirac has interpreted the negative energy states.
4. Develop Diracs relativistic wave equation for a free particle and obtain its plain
wave solution.
5. Deduce relativistic Klein Gordon equation. Can we use this equation for pions and
kaons. Show that its probability expression is indefinite.
6. Establish the Diracs relativistic wave equation for an electron and calculate its
magnetic moment.
7. Obtain the eigenvectors for Diracs relativistic wave equation of a free particle.
8. Develop the relativistic wave equation for the particle of spin zero. Discuss the
difficulties involved in interpreting the probability density. Derive the expression
charge and current densities from the continuity equation.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
91
9. Prove the following properties of Dirac matrices on the basis of Diracs theory
a. Square of matrices are unity
b. Eigenvlues of the matrices are 1.
c. Trace of the matrices are zero.
d. Dimension of the matrices are of even order.
e. The matrices anticommuting among themselves.
9.9. References
1. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan. Tata
McGraw Hill Publishers.
2. Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
3. Quantum Mechanics V. Devanathan Narosa Publishing house.
4. Quantum Mechanics S.L. Gupta, V. Kumar, H.V. Sharma, R.C. Sharma.
Jai Prakash Nath & Co Ltd.
5. Quantum Mechanics Satya Prakash Sultan Chand Publishers .
6. Angular Momentum Techniques in Quantum Mechanics V. Devanathan
Kluwer Academic Publishers.
7. Quantum Mechanics John L. Powell and Bernd Crasemann Narosa
Publishing house.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
92
LESSON 10
SPIN AND SPIN MAGNETIC MOMENT OF ELECTRON
= x y , a
p - x py .yp
x p x , ya
Here we have used the fact that x does not commute with px and y cannot
commute with py but they commute will all other quantities.
Also [ x, x p] = ,yy =p . ih
[ Lz , H] x= [x ,y p] p , yy
-x a px a py
= ih pxa
- x pxa y (1)
1 s z 0
to be where
2 z
= 0
z
s z
(4)
Then (Lz + Sz) commutes with H
This watermark does not appear in the registered version - http://www.clicktoconvert.com
93
is 0
= -i z ,
0 is z
s 0
= z
0 sz
= z (5)
Now [S z , H ] = 1 [ z ,H ],
2
i
=-
2
[
a xa y , H , ]
i
=-
2
[
a xa y , a x p x + a y p y + a z p z + b m , ]
i
=
2
{
a-x ay, xa
p+ x a ,x ayp a
y , y }
i
=
2
{
(a-x ay ax -ax ax )yap+ xa( x a y-a y apy ax ay ) y },
i
=
2
{(
- a x2 -
ay )
ax 2- yap x a
+2 ( x a y+ a2px ay ) },
y
i
=- {- 2a y p x + 2a x p y },
2
= -i{- a y p x + a x p y },
= i {a
- x p y-a y p x }, (6)
[(Lz + S z ), H ] = 0 = i d (Lz + S z ).
dt
Therefore we find that the angular momentum is conserved only if we include the spin
angular momentum operator
1 1 s 0
Sz = = z (7)
2 z
2 0 s z
This watermark does not appear in the registered version - http://www.clicktoconvert.com
94
1 0
= (8)
0 1
1
We find that the eigenvalue of = 1. Therefore the eigenvalue of S
z z = ( 1) are
2
1 1
+ and - .From these facts it is evident that the Dirac equation describes particles
2 2
1
of Spin such as electrons.
2
where A and j are vector and scalar potentials respectively. Therefore, the Dirac
equation becomes,
u r
u r r e A 2
(E - ej) y = ca - p b - + y mc
c
u r
u r r e A 2
ih - ej y = ca - p b - + y mc
t c
( )
y r, t = y (r ) e h
This watermark does not appear in the registered version - http://www.clicktoconvert.com
95
( )
y r, t = y ()r
- iN .E
h
e
R t
- i m2 c t
h
e
r - i m2 c t
= y' ( rt ) , e h
( )
wherey ' r, t = y (r ) e
-iE N.R t
h
is the non-relativistic wave form
y r - i m2 c - i m2 c
t
ih = ih yt ' y+'
(r,)
h
e
t t h
- imc 2
2
t
= ih + mc e h
y'
t
Substituting in the Dirac equation
u r
- imc 2
t u r u re A 2
- i m2 c
t
2 h
ih + mc - ej y '.e h = ca . p - b + mc y ' e
t c
Replacing the wave function y as a two component wave function y A and y B (large
&
small components)
u r u r
y A 0 s u 1r e A 0 y A
2
ih + mc - ej = c u r p. - + mc2 y
t y B s 0
c 0 - 1
B
u r
2 u r u r e A 2
ih + mc - ej y A = cs . p - y B +y A mc
t c
u r
y A u r u r e A
ih = cs . p -y f B + y eA (1)
t c
Similarly
u r
2 u r u r e A 2
ih + mc - ej y B = cs . p - y A y
- Bmc
t c
u r
2 u r u r e A
ih + 2mc - ej y B = cs . p -y A (2)
t c
Since
y A y B
ih = ih = ih {y ' (r, } t)
t t t
This watermark does not appear in the registered version - http://www.clicktoconvert.com
96
-iE N.R t
= ih y (r) e h
t
- iN .ER t
E
= ih -i N .R y h
e
h
-iE N.R t
h
= E N .R y e
Equation (2) becomes,
u r
u r u r e A
cs . p -y
{E N.R }
+ 2mc 2 - ej y B = c
A
u r
u r u r e A
cs . p y- A
yB = c
{ EN . R + 2mc2
f- } e
u r
y A u r u r e A 1
ih = cs . p - 2
t c +
EN . R2mc - f e
u r
u r u r e A
-cs . y-p + f Ae yA
c
Since EN.R <<< 2mc2 & assuming the magnetic field to be constant, ie j = 0,
EN.R + 2mc2 2mc2
u r u r
y A u r eu r A u 1r u e r A
ih = cs . p- s c 2 p y . - A
t c 2mc
c
u r u r
1 ue r uAr u er u rA
= s . p - s p . - (3)
2m c c
R.H.S can be reduced by the following relation
urururur u r u r u r u r u r
a .B a ( ) ( ).C = B . C+ a Bi . C
u r
where B & C commutes with a but not necessarily with each other. Here
u r
u r u r u re A
B = C=p - necessarily with each other. Now,
c
eA eA eA eA
s . p - s. p - = p - eA . p - eA + is p - p-
c c
c c c c
This watermark does not appear in the registered version - http://www.clicktoconvert.com
97
e u r u r u 2er u r 2
p2
c
(
- p .+ pA + A.A 2
c
)
= u r -u re u r u r u r
+ s i . +p . A
c
(p A. )
e2 2 - e
= p2 +
c 2
A
c
{ } - e
c
2 A. p - ih . A + is . (-ih A)
Equation (3) becomes
p2 2
e e urur
+ A2 A - p .
y A 2m 2 m 2
c m c
\ ih =
t i h ue r uh er u r u r u r
+ (
A . -s A .
2 m c 2 m c
)
Since A = B , the external magnetic field,
p2 2
e 2 e
urur
+ A A - p .
y A 2m 2 m 2
c m c
\ ih =
t i h ue r uh er u r u r
+ A. B-s .
2 m c 2 m c
This is nothing but, Schrdinger equation in N. R approximation.
The last term in the above expression gives the interaction of electron spin,
()
with the external applied Magnetic field B . Thus we find that the electrons behaves
eh
as a particle having a spin magnetic moment s where the spin of the e is
2mc
1
S= hs
2
Now it can be shown that
A =
1
2
( )
B r
Then
e
mc
A. p =
1 e
2 mc
( )
B r p
1 e
= B.r p
2 mc
e u r r r
B . L (where L = r ispthe orbital angular
= mc
momentum)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
98
y A p2 2
e 2 e
u i r u r ehu r u ehr u r u r
\ ih = + A + - B . L A . Bs
- .
t 2m 2 m 2c 2 m c 2 m c 2 m c
p2 2
e i he u r eu r u r u r u r
= +
2m 2 m c 22
A2+ A. - B
m c 2 m c
.+ L h s ( )
1
Q S hs
2
y A p2 2
e i he u r ehu r u r u r u r
ih
t
= +
2m 2 m c 22
A2+ . -
A
m c 2 m c
B .+ L S 2( )
The last term gives the interaction of the orbital and spin angular momenta
10.6. References
1. Angular Momentum Techniques in Quantum Mechanics V. Devanathan
Kluwer Academic Publishers.
2. Quantum Mechanics V. Devanathan Narosa Publishing house.
3. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan. Tata
McGraw Hill Publishers.
4. A text book of Quantum Mechanics S.L. Kakani Kedarnath Publishers.
Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
5. Quantum Mechanics A.K. Ghatak and S. Lokanathan McMillan Publishers.
6. Quantum Mechanics Chatwal Anand Himlayan Publishing house.
7. Quantum Mechanics John L. Powell and Bernd Crasemann Narosa
Publishing house.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
99
LESSON 11
SCATTERING THEORY
11.2. INTRODUCTION
When a beam of particles of any kind is directed at matter the particles get
deflected from their original path since it collides with particles of matter. The
scattering experiments provide useful information about the nature forces, and
interaction between the particles of matter. In some collisions, the energy of the
incident particle does not change. Such collisions are called elastic collisions or
otherwise it is called inelastic collision.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
100
Scattering Plane
Reference Plane
f
Target q
Incident beam
Co
ne
ofS
cat
teri
ng
of
an
gle
q
and f indicates the direction of scattered particle with z-axis as the polar axis. Here q
is the angle between the incident and scattered direction and is called as the angle of
scattering. The azimuthal angle f specifies the orientation of this scattering plane with
respect to some reference plane.
101
a Dt
aI
ds
DN = d W t D I (1)
dW
ds
where the proportionality constant is called the differential scattering cross
dW
section. Therefore the differential scattering CROSS SECTION. is defined as the
Number. of particles scattered per unit solid angle, per unit time, per unit incident
flux. It has the dimension of area.
ds ND
\ =
dW dW t D I
The total scattering cross section is obtained by integrating the differential
cross section over all the polar angles q and f
p 2p
ds
s = q s qi nd fd
0 0
dW
102
Where f(q) is called the scattering amplitude. The incident flux is given by
2
I i n =cUi n c J
2
= el k
z
J
= 1 J
= 1 J
ds
DN = d W t D I (B)
dW
Comparing equations A & B.
ds 2
= f (q )
dW
Thus we find that the differential cross section is equal to the square of scattering
amplitude. Thus the scattering amplitude is related to the experimentally observable
ds
differential scattering cross section. Since has the dimension of area (m2 ), f() has
dW
the dimension of length (m).
This watermark does not appear in the registered version - http://www.clicktoconvert.com
103
11.9. References
104
LESSON 12
SCATTERING THEORY
12.2.1. INTRODUCTION
The method of partial wave is elegant procedure for the analysis of elastic
scattering and low energy scattering. It is done in two steps. First wave function u(r,
) which represents the sum of incident and scattered wave is obtained in terms of
partial waves. Secondly the asymptotic value of this wave function is equated to the
wave function u(r, ) given by
r r
r f (q e)i
i k z
.k r
u ( q,r =) e + r (1)
r
A plane wave ei k zcan be expanded as a linear combination of spherical waves using
Rayleighs scattering formula.
ei =kiz l l (+j 2 1K )l (r )l P( c oq s ) (2)
l =0
105
s i k n-r p(l2 )
jl ( K )r=
k r
i ( K - rp 2) l -i (Kp 2- r )l
e e-
=
2i k r
Substituting in equation (2)
1z l i
ei =k i (l k
2+r e(1l -/ pi-)2 -ek) ( (pr - / cl 2P )o l s q)
2i k l =r0
This form shows that each partial wave can be represented as the sum of
incoming and outgoing spherical wave. Partial wave methods can be applied when the
potential is central and of finite range. This method is really useful whenever the
energy of the incident particle is low. Here the incident beam of particles a r e
represented by a plane wave with definite linear momentum with no definite angular
momentum. Hence a plane wave can be decomposed into infinite number. o f
components each of which corresponds to definite angular momentum (i.e) into
partial waves. The process by which a plane wave is decomposed into a partial wave
is called the partial wave analysis.
For a spherically symmetric potential having asymptotic behaviour, the total
wave function will consist of the incident plane wave and the outgoing spherical
wave.
f (q e)i k r
u ( q,r =) ei k +z (3)
r
Which is independent of . Using Rayleighs expansion for plane wave
ei =kiz l l (+j 2 1K )l (r )l P( c oq s ) (4)
l =0
s i Kn -r( p/l 2 )
But jl ( K )r=
K r
Expressing in terms of exponential.
K-/ pr 2 l -) i (K
p - /r
ei ( e- 2l )
jl ( K )r= (3)
2i k r
il ( + lK2( -/ r-p 2 e-l1) (pi - / e2k) ) r f(l
i i
)ek q r
u ( q =, )r ( cq P o + sl ) (4)
l= o 2i k r r
The Schrdinger equation for the scattering problem can be written as
2 2m
+( E ) V-( u , rq=)
h 2
0
This watermark does not appear in the registered version - http://www.clicktoconvert.com
106
The Schrdinger equation can be separated into radial part and angular part.
The angular part of the solution will be spherical harmonics., Ylm (, ) which will
crumble to the legendre polynomial Pl ( c oq s for
) spherically symmetric potential
which is independent of .
The radial part of the solution Rl (r) will involve spherical Bessel function and
spherical Neumann function in the field free case. Therefore the total wave function
can be written as
u( q
,r )= R( )lr ( qlPc o s ) (5)
l =0
2m E 2 mV
Where K 2 = U2 , =(r )
h h2
The solution of the above equation (i.e) the radial wave function will be a linear
combination of jl ( K )rand hl ( K )r
- c o K-s r( p/l 2 )
Where hl ( K )r=
K r
However it should be noted that jl ( K )r is a well behaved function at r=0
where hl ( K )r is not so, since it becomes infinite at r=0. Therefore we choose the
radial wave function as R l ( r proportional
) to
Rl ( )ar c jlod lsK ( d l)-r ls hK
i n r( )
In this expression if we choose d l =0 for the field free case, then at origin, only
jl ( K )r will remain. At all other points both jl ( K )r and hl ( k rwill
) be present. This is
107
s iK n- l[p r/ 2 ]K - c lpo r s ( / 2 )
= Al c do s l+P s d i l n l c( q o s )
l =0 K r K r
1
= Al [ c dolKs s -i nrpl( / l2+K) sd i lnr -c ol psP( / 2 )cq] ( o s )
l =0 K r
1
= Al . s Ki- n lp
r(l +/ P2 dl ) (c qo s )
K lr- 0
1 -Kl/p
= l
A i(
e -r+ 2l )di e-(lKl p(-/ rl P2d+o cs q) (6)
2i K l =r0
1 l 1 l p / 2 2 id l
i ( + 2l i
1l P) ql+c -e po/ 2 sqf ( ) = i +( l i2 e1l - )e c P lo sq
2i kl = 0 2i kl = 0
1 l l 1 l
f( q ) = i ( l +2e 1- i ep /) 2 2cidP-
l
lo sq i ( 2l+ 1i ) P l
l c oqes- p/ 2
2i kl = 0 2i kl = 0
1 l
= i ( l 2+ el1 - i)p /Pc2 l o- qes 2(id l 1 )
2i kl = 0
= eid l [ e-
idl
eid-l]
= eid l . 2i s idnl
1 i p / l 2 i p / -l 2
f( q ) = e + ( l 2 1e ) cP lo seq 2idl si di nl
2i kl = 0
1
f( q ) = (+l2 1 P)l c oqe sldll s di n
k l =0
(7)
This watermark does not appear in the registered version - http://www.clicktoconvert.com
108
1 - ld
= +( l2 1 +l) '( 2P 1l ) Pcq o' se lcq oldel s sdl ' i' ln lsd i n
K2 l l '
2p 2
s= +( l2d 1 l)+ ( 2e ' id1lde )-idl s i n l s i dl n' ' l l
k2 l 'l 2l + 1
2p 2 2 2
s=
k2
+( 2l d1 ) s i n
l =0
l
2l + 1
4p 2
s=
k2
+( 2ld1 ) s i n
l =0
l (8)
A q = 0t c o s =l(P 1 =) 1 for
lq
P , all values of l.
1
f ( 0= )l( + 2 1e ) i dsl i dnl
k l=
109
4p 1
= . (+2 1l ) s i 2nd l (2)
k kl= 0
scattering amplitude for q = 0. In other words the imaginary part of forward scattering
amplitude measures the intensity which the incident beam suffers because of
scattering.
12.4.1 PHASESHIFT
The effect of introducing the potential into the picture is just to change the
phase angle of the normally outgoing scattered wave by an angle d l . We have just
seen that the radial wave function in the absence of potential is proportional to
s i Kn -r(lp/ 2 )
jl (K )r=
k r
Where as in the presence of potential it is proportional to
s i Kn -r(lp/ 2 )
jl (K )r= l + d
k r
Now in the expression for local wave number the radial Schrdingers wave equation
d2 u l ( +1 )
+ [k 2 ( u- 2 ) r ]+l (u =) r 0
d 2 r r
We find that for an attracting potential u(r) is negative and have k 2 + u(r) increases or
the wavelength decreases.
For an attractive potential the waves are pulled in and for a repulsive potential
u(r) is positive and the wavelength of the scattered wave increases and a repulsive
potential the waves are pushed out.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
110
It is to be noted that a phase shift will be introduced only with in the potential of will
be maintained through out.
d l= 0
d l= +ve
d l= ve
RESULTS : -
111
2 m E 22m
letting = k 2, V ( =r ) U( r)
h2 h
2 u ( r +)2 k (u =)r U
( )r u( r)
( 2 ) k2+( u ) r= (U )r (u r) (1)
Noting that ( 2 +k2) e lk 2 is zero, we can write the solution for Schrdinger equation as
U(r) = ei k G+z r ( -r ' U) (r ' u) r( d' )r ' (2)
This can be easily verified by operating ( 2 +k2) on both the sides of equation (2)
( 2) k(2+ )u =0r (r +' rd ) -U( r' )u (r ' d ) r
( 2 ) k2+( u ) r= (U )r (u r)
Thus we find that the expression for u(r) given by equation (2) is valid.
Now denoting uo ( r =) ei k ,z we have
u(r) = uo (r )+G( r '-r) U( r' )u (r 'd ) r ' (3)
The above series can be extended further by replacing r by r by r ' b' y' r' &' so on.
Equation (4) becomes
u ( 'r ' =)u (o 'r ' +)G( ' r' -' r' ' U
) ( 'r ' 'u) (r ' 'd' ) r' ' '
+ G( r'- )r ( U' ) r-( G' ' r' ' r' u)( r' ' G) ( r' -'r ' U
' ' )r ( u' ' r' )d( r' ' d' )r' 'd' 'r ' '
Such a series is called Born series or Neumanns series. This expression is evidently a
perturbation expansion of U(r) with the potential treated as perturbation. If the
incident energy is very large, then the series will be converging rapidly. In such a
case, only the first few terms will contribute and it will be sufficient to terminate the
series after two or three terms. Thus for first Born approximation, we will have two
terms.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
112
Therefore in the first Born approximation the wave function will be.
u ( r )= u0( r)+ G( r' - )r0 U
( ' r )u ( r ' d ) r '
i k z 1 ei ( k - ko) r
=e U - r u - ( r ' ) d ( ' r ) '
4p r
-1 - i ( Ko- K
)i r
f B (q , f=) e
4p
Defining K o - Kl = K the momentum transfer function.
-1
Ko
K = ko -ki
ki
From equation (7), it is clear that for the first amplitude depends only on the
momentum transfer and not on the initial and final momenta and the angle of
scattering. For a spherically symmetric potential, the potential will be independent of
angle f and therefore.
p 2p
-1
f B (q e , = U)f r( - i)rqk r
sd 2
qi r nd f d
4p 0 0 0
This watermark does not appear in the registered version - http://www.clicktoconvert.com
113
p
-1
= e - i kUr ( r ) 2r s di rqn dq
2 0 0
p
-1 2 ( c o i k -sr q)
= r U( r) -d ( r c e od s q)
2 0 0
p
-1 ( c- eoi k r s q )
= r 2 U( r )d r
2 0 i k r 0
-1 2 - ei k r ei- k r
= r U( r )d r
2 0
i k r
-1 ei k -r ei- k r
f B (qr =
) U ( r ) ds ri kn r since s i kn =r
k 0 2i
-1 ei (k -r' r) m
V ( r ' ei k' )dz r'
2p r -' r h 2
=
114
1 2
2 m 2 i
k r i k rw
g( e=0 r2) 4 Vp( r2 d) r(w
e) d.
4p h 0 1 -
2
1
m2 i k r
i k r i k rw
=
h4 e e V r( )r . d r e wd
0 0 1 -
1 2
i k w
r
r e
m2
i k r
= 4
h e V (r )r . d
0
i k r
-1
2
ei -e
k r i- k r
2
m
i k r
= 4 e V (r )r . d r
h 0 i k r
2
m2
= 2 4
e 2i k)er 0 V
(- . (r d) r
k h 0
2
2 m2
g( 0e= 2) 4V( - 2i1rk r ) <d . (r < ) 1
k h 0
115
12.8 References
1. Quantum mechanics K. Sankar Narosa publishing house
2. Quantum Mechanics A.K. Ghatak and S. Lokanathan McMillan
Publishers.
3. Introduction to Quantum Mechanics David J. Griffiths, Prentice Hall.
4. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan.
Tata McGraw Hill Publishers.
5. Basic Quantum Mechanics Ajoy. Ghatak Tata McGraw Hill Publishers
6. Introductory Quantum Mechanics Richard L. Liboff Addition - Willey.
7. Quantum Mechanics John L. Powell and Bernd Crasemann Narosa
Publishing house.
8. Modern Quantum Mechanics J.J. Sakurai.
9. http://thequantumexchange.org/
This watermark does not appear in the registered version - http://www.clicktoconvert.com
116
LESSON 13
APPLICATIONS OF SCATTERING THEORY
ze 2
V (r ) = - (1)
r
1
f (q ) = - r U (r ) sin kr dr
k 0
2 m 2 sin kr
= ze dr
h2k 0
r
2 mze 2 1
= .
h2k k
2mze 2
= 2 2 . (2)
h k
q
We know that momentum transfer k = 2k sin
2
2 mze 2 mze 2
\ f (q ) = = (3)
2 q
2 2 2 2 q 2
h 4k sin 2h k sin
2 2
117
ds 2 m 2 z 2e4
= f k (W) =
dW q
4h 4 k 4 sin 4
2
m 2 z 2e4
, where p = hk
= (4)
q 4 4
4 p sin
2
This coincides with the classical Rutherford formula for coulomb scattering.
f B (q e f,= U ) -ri k( r d ) r
4p
-1 m2 - i k r
4p h 2
= e V r( d) r
-1 mVo - i k re -a .r
2p h 2
= e d r
r
a p2 p
-1 mVo - i k cr q o e -as.r 2
=
2p h 2 e
0 0 0
r d
r
r s idqn qd f
a p
-1 mVo
= 2p e -(a rr i-
dc
k q-r ( c o s )
r- qo e ds )
2p h 2 0 0
m vo -a r e -i k(-rc oq s )
= e r d r -i k r
h2 0 0
-i k r
mV ei k -r e
= 2o e -a dr r
h 0 i k r
-2mVo -a r
=
h 2 k e
0
ks ir n d r
118
Here q is the angle between the incident wave vector k and the unit vector r in the
scattered direction. From the above expression, it is clear that, the scattering
amplitude is independent of angle f . This independence is a general result for elastic
scattering by a central potential.
The differential scattering cross section can be written as
ds 2
= f (q ,f
dW
4m 2V 2 o 1
=
h 4 (a 2
4+ s2 k i q n2 / 22 )
Thus , we can see that, the differential scattering cross section varies with k and so it
also varies with energy E. In the limit of zero energy (k 0), the differential cross
section is
4m 2V 2 o
s=
h 4a 4
Thus the differential cross section is independent of q as well as of f . Except at zero
energy, s peaks in the forward direction ( q =0) and decreases monotonically as q
varies from 0 to p , which is shown is fig. below. It is to be noted that, with in the
Born Approx, the differential cross section is independent of sign of Vo and gives the
same result if the yukawa potential is attractive (V0 <0 ) or repulsive (V0 >0 )
1.0
0.9
(s / )
4m2 v 2 ao4 h4
0.1
0.8
0.2
0.7
0 p
q
This watermark does not appear in the registered version - http://www.clicktoconvert.com
119
The starting point of calculations on all except the lightest atoms is the central
field approximation. This assumes that each of the atomic electrons moves in a
spherically, symmetric potential energy V(r) that is produced by the nuclear and all
the other electrons. The approximation is a good one if the deviation from the V(r) for
one election produced by close passage of other electrons is relatively small. This is
actually the case, since the constant nuclear potential is of the order of Z times as
large as the fluctuating potential due to each nearby electron, and the latter varies
quite slowly (inversely) with the separation distance. The two principal problems are
the correction of the approximate results obtained from it. Before considering these
problems, we discuss some general properties of the central field.
The potential energy V(r) for a neutral atom has the coulomb form c2 /r at a
great distance r from the nuclear, since the removal of the electron whose potential is
being measured leaves a singly charged positive ion. The electron in the hydrogen
atom, for which the potential energy is c2 /r at all r has an infinite number of bound
energy levels characterized by the quantum numbers n, l and m. An infinite number of
energy levels is also expected for V(r), since for large n the electron wave function is
small near the nucleus and only the form of V(r) of large r is significant .An important
difference between the two situations is the degeneracy between states of the same n
and different L that occurs in hydrogen is removed in a noncoulomb central field. This
is because the electrons that have smaller angular momentum penetrate closer to the
nucleus, and V(r) is stronger (more negative) than c2 /r there, since the nucleus is less
completely screened by the other electrons. Thus for given n, the states of lowest l
have the lowest energy. The degeneracy with respect to m is not affected, since this
occurs whenever the potential is symmetric.
Because of the spin, four quantum numbers n, l , me and ms are required to
specify the state of an electron in a central field. The orbital quantum numbers l and
1
ml are the same as l and m in the hydrogen atom, ms = specifies the spin
2
orientation and n is a natural generalization of the total quantum number that appears
in hydrogen.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
120
Two methods have been used for the determination of the potential energy
V(r). The first of these Thomas. Fermi statistical model assumes that V(r) varies
slowly enough in an electron wavelength so that many electrons can be localized
within a volume over which the potential charges by a small fraction of itself. The
electrons can then be treated by statistical mechanics and obey the Femi Dirac
statistics. At normal temperatures the thermal energy kT is very small in comparison
with V(r) every where except at the edge of atom where the chance of finding an
electron is small. In this case, the Fermi Dirac statistics requires that the electron
states fill in order of increasing energy, as assumed.
The number of electron states in a cube if edge length L at the walls of which
the wave functions obey periodic boundary conditions was computed to be
3
L
dk x dk y dk 2 . This must be multiplied by 2 to take account of the two
2p
possible spin states; then the number of states for which the momentum P = hk
is less than or equal to P0 .
p0
3 h p 2p 3
L 2 Po L3
2 d 0 k dk sin q dq df =
2p 0 0 3p 2 h 3
If all these states are occupied, the number of electrons per unit volume whose
2 3
Po Po
kinetic energy does not exceed is
2m 3p 2 h 3
Now the maximum kinetic energy at any distance r form the nucleus is V(r), since
otherwise the electrons would escape the atom. We thus obtain a relation between the
volume density of electrons, n(r) and the pontentialenergy :
n(r) =
[- 2m v(r )]3 2 (1)
3p 2 h 3
This watermark does not appear in the registered version - http://www.clicktoconvert.com
121
Elimination of n(r) from (1) and (5) leads to an equation. for V(r).
3
1 d ( d2 -V) 4e 2 [ -2m V
( r)] 2
r = (3)
r 2 d r d r 3
3p h
we put
- z 2e
V(r) = c r =b x
r
2
1 3p 3
h2 0 . 8 8ao5
b= = 1 1
(4)
2 4 m 2
e3 z 3
z
122
1
2
d 2c 3
2
x =c
dx 2
c = 1 at x = 0 c = 0 at x = (5)
The most accurate solution of Equation (5) was computed by Bush and
Caldwell with the help of the original differential analyzer and is expressed form of
numerical table.
Equation (4) show that the radius of an atom is inversely proportional to
the cube root of atomic number, if this radius is interpreted to be that of a sphere that
enclose a fixed fraction of all the electrons. These equations can also be used to show
that the Thomas Fermi approximate improves with increasing z. The potential at the
atomic radius is proportional to z 1/3 so that a typical electron wavelength is
proportional to z 2/3 . The distance over which the potential changes by a definite
fraction of itself is proportional to the atomic radices or z -1/3. Thus the fraction change
of the potential in an electron wavelength is proportional to z -1/3 and decreases with z.
Moreover since the number of electrons is equal to z, the use of the statistical method
is better justified s z increases.
- h2 2 ze 2 2 e2 3
k - + | u i (r j ) | d r j u k (rk ) = k uk (rk ) (1)
2m rk jk rj k
where rjk = |rj-rk |. If there are z electrons in the atom, equation (1) constitutes a set of z
simultaneous nonlinear integral differential equations for the z functions uk (rk ). It is
therefore not feasible to solve these equations. directly, and Hartree used a method of
successive approximation.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
123
where j > k implies a double summation over all different pairs of indices j and k we
wish to minimize the expectation value of H,
From (2) and (3) we obtain
3 3
. . . . . . y. . . H*y. . . .d. . . . r. . . . d. . . . .r.
1 2
* -h 2 2 z e
= u k ( rk ) -u2k (kr
3
d)
k k r
k 2m r k
e2
+ > u *j( r k )u k (r* ku) jr( 3 uj ) rk ( dk3) r j d k r (4)
j k rj k
This watermark does not appear in the registered version - http://www.clicktoconvert.com
124
since the uk are normalized. The optimum y is obtained by varying each of the uk
separately to minimize equation (4). The only dependence of equation (4) on a
particular one electron function uk is through the terms.
* -h 2 2 2 z e 3
uk ( rk ) uk -(kr d)
k r
2m r k
e2
+ u *j( r j )u k (r* ) j r( 3 ju) rk ( dk 3) r j d k r
ku
j k rj k
= u k* ( rk ) H k u k ( rk ) d 3 rk (5)
-h 2 2 z 2 2 e 2
e
Hk k - +j | u k( )r3 | dj r
2m rk j kr j k
The integral in equation (5) is the expectation value of the operation Hk for the
function uk . This is a minimum when uk is an eigenfunction of Hk that corresponds to
its lowest eigenvalue k .
Hk uk = k uk (6)
Since equations (1) and (6) are identical, we see that the Hartree wave
functions are the best from the point of view of the variation method that can be
written in the form of equation (2)
The energy associated with this wave function is just the integral in equation
(4) which can be written with the help of equation (1)
3
..................... u * H y d r1 ................... d 3 r2 = k
k
e2
- > | u j (r j ) | 2 | u k (rk ) | 2 d 3 r j d 3 rk
j k r jk
The electrostatic interaction terms between electrons are counted twice in the
summation over k and so have to be subtracted to give (7). Thus the energy of the
atom is not just the sum of the k , although each k . is roughly the energy of removal
of the kth electron. This last is not strictly true, since the removal of an electron alters
the self consistent fields and hence the wave fr. s for the remaining electrons.
However, k is found to be an especially good approximation to the energy of
removal in the case of an inner electron (x ray level).
This watermark does not appear in the registered version - http://www.clicktoconvert.com
125
13.9 References
1. Quantum Mechanics A.K. Ghatak and S. Lokanathan McMillan
2. Publishers.
3. Introduction to Quantum Mechanics David J. Griffiths, Prentice Hall.
4. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan.
5. Tata McGraw Hill Publishers.
6. Basic Quantum Mechanics Ajoy. Ghatak Tata McGraw Hill Publishers
7. Introductory Quantum Mechanics Richard L. Liboff Addition - Willey.
8. Quantum Mechanics John L. Powell and Bernd Crasemann Narosa
9. Publishing house.
10. Modern Quantum Mechanics J.J. Sakurai.
11. http://thequantumexchange.org/
This watermark does not appear in the registered version - http://www.clicktoconvert.com
126
LESSON 14
APPLICATION OF SCATTERING THEORY TO
ATOMIC AND MOLECULAR STRUCTURES
127
The difference in energy between states of different j is due to the L.S term in
(1) and can be found from its expectation value or diagonal matrix element.
2 u r r
J 2 = LS
( + 2 ) L= 2 + S + 2 L.S (2)
since e, j and s are all good quantum number (s = ). Equation (2) can be solved for
the diagonal matrix element of L.S.
u r r 3
lL.Sl
j j = [j +(j1) +l (l1) -J 2k (3)
4
1 l nl 1
if j = l +
2 2 (4)
-1 1
(l + 1) nl if j = l -
2 2
nl = h 2 |R nl (r ) | 2 (r ) r 2 dr l >0
0
where Rnl (r) is the normalized radial part of the unperturbed eigenfunction associated
with the nl configuration. Since V(r) represents an attractive potential energy, (r) is
positive and nl is positive. Thus equation (4) shows that the state with higher j has
the higher energy. The pair of states is called a doublet the doublet structure
characterizes all the moderately excited levels of the alkali atoms except those for
which l = 0 in which case j can only be .
The doublet separations can be calculated from equation (4) if radial function
is known using the generating function for the associated lauguerre polynomials.
lj
h2 z 2 e 1 2
nl = R n r(l d) r
2 m2 2 c 0
r
This watermark does not appear in the registered version - http://www.clicktoconvert.com
128
e2h 2 Z 4
= (5)
1
2 m 2 c 2 a o 3 n 3 l l + (L + 1)
2
This is valid only for l >0 ; the singularity in (r) at r 0 makes the integral for no
diverge there, so that the perturbation approximation is not valid. It follows from (4)
and (5) that the doublet separation is proportional to n-3 and this in fair agreement with
observation. The absolute value of the doublet separation and its dependence on l are
not given at all by this simple theory, since the effective z is difficult to estimate and
depends marked by on l because of penetration.
We now calculate the intensities of the two lines of the allowed doublet 2P
3 1 1 1
2S & 2P 2S under the assumption that the radial wave functions
2 2 2 2
are the same for the two excited 2P states. Transitions of this type give rise to the
principal series in the alkali spectra. From the equation.
2
4 e 2 k 2 wkn 2 4 e 2 wkn
| < k | r | n >| = | < k | r | n >| 2
3h c 3h c3
the spontaneous transition probabilities and hence the observed intensities if the two p
states are equally likely to be occupied, are proportional to the squares of the dipole
matrix elements
The dependence of the two excited 2p states and the ground 2S state as the
angular and spin co-ordinates of the electron is obtained by finding linear combination
of products of the four spherical harmonics y1 , 1(q, f), y1 , 0 (q, f), y1,-1 (q, f), y0 ,0 (q, f)
and the two spin wave functions (+) and (-) that are eigenfunctions of J2 and Jz. These
combinations can be obtained from the Clebsch Gordan co-efficient.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
129
3
m = y+ ( ) 1 1,
2
-1
1 1
= + 3 2 +22 ( y1)-, 0( y1 ), 1
2
-1
-1 1
2 p3 = 3 1 2-, y (2 0 +2 ) 1 y( +, )1
2
2
-3
= - y1 , 1
2
1-1
1 22
m = - y 3 + ( )+ 1 y2, 0-( 1 ), 1
2
-1 1
-1 2
2
2 p 1 =3 1 ( -,y ) 0 -2 1 y( +, ) 1 -
2
2
1
2s1 = 0(m , +)y 0
2
2
-1
= - ( )y0 , 0
2
The wave functions equation (6) can be used to calculate the matrix elements of
x = r sinq cosf y = r sinq sinf and z = r cos q. We assume that the radial functions
associated with equation (6) are all the Same, so that the radial part of the matrix
element integral is a common factor throughout. The angle parts of the integrals are
easily evaluated by making use of the explicit expressions for the ys in terms of q and
f. The products of spin fns, follows the simple rules (+) (+) = 1, (-) (+) = 0 etc., In this
way we obtain the following values for the squares of the magnitudes of the indicated
1
matrix elements, expressed in units of of the common radial factor :
18
3 1
m = t =o m = | x 2| =| y2| 3 z2=| | 0
2 2
3 1
= t= o - x =| 2|y |=2 | z 2| = | 0
2 2
1 1 2
2 p 3 2 1 s | =t | o| = | x 0 = y 2| |2
z= 4
2 2
2 2
1 -1
= t =o | x 2|= | y 2| 1 =z2 | | 0
2 2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
130
1 1
m = to m = | x |2 = | y |2 = 0 | z |2 = 2
2 2
1 -1
= to | x |2 = | y |2 = 2 | z |2 = 0
2 2
-1 -3
Similar results are obtained for the transitions that start form m = and .
2 2
It follows from equation (7) that the sum of the intensities of all the lines that
originate on each of the four 2P3/2 states is equal to 6, in the above units. It is
expected that these sums are equal since the four values of m differ only in the
orientation of the angular momentum, and this should not affect the intensity.
However the total intensity form each of the two 2P1/2 states is also equal to 6. The
equality of total intensities from each state formed form a given L and S is a general
property of LS coupling; this makes the absolved intensity, which is that from all the
states that are degenerate with respect to m, proportional to 2J+1. In the example
considered here, the two lines for the doublet have intensities in the ratio 2:1. This is
observed for the lowest doublets of the alkalis, although for the higher doublets the
intensity ratio exceeds 2. This is because the spin orbit energy actually mixes
different configurations (2p states with same j but different n ) ; the amount of mixing
is different for the two j values, so that the two radial functions are not the same. A
small admixture of the low- intensity upper states in the high intensity lowest 2p
states has little effect, whereas in the opposite case there is a large effect on the
doublet intensity ratio.
131
in the case of the hydrogen molecule in the magnitude R of the distance between the
two hydrogen nuclei. The Hamiltonian is that given in equation
H = H0 + H '
h2 2
e2 e
H 0 = -1 ( 2 2+ 2) - - (2)
2m r 1r 2
e 2 e2 e2 2
e
H'= + - -
R 1r 2 rB1 r2A
132
The correction of the approximate results obtained from the central field. Two
terms are omitted in the central field approximation: the difference between the
actual electrostatic interaction between electrons and the average interaction that is
included in the central field, and the spin orbit energy. The latte is an interaction
energy between the spin and the orbital motion of each electron and has the form
u r u r
x (r
k
k )Lk .Sk (1)
u r u r r
here, L k is the orbital angular momentum operator rk x k ofpthe kth electron and has
u r u2r
the properties of the J operator. The eigenvalues of L k and Lkz are given in terms of
the quantum numbers l and ml for the electron as l ( +l1 h)2 and ml h , respectively.
r 1
Sk is the spin angular momentum hs k of the kth electron. The function x ( r is
) given
2
by
1 1d V
x (r =
)
2m 2 c2 r d r
in terms of central field potential V (r).
In considering the effect of these terms, we shall assume that the perturbed
eigenfucntions, which are linear combinations of various configuration wave
functions, have only negligibly small amounts of all but one configuration mixed in
them. From perturbation theory it is apparent that this is the case if the inter
configuration matrix elements of the perturbation are small in comparison with the
energy intervals between unperturbed configuration energies.
It can be shown that from equation (1) that includes electron in full shells is
zero, since the function x ( r is
) the same for all electrons in the shell and contributions
from electrons with opposite ml and ms cancel. Thus the electrons in the full shell can
be ignored and the summation extended only over the remaining electrons. The case
in which there is just one electron outside the full shells is of interest in connection
with the ground state and low excited states of alkali atoms.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
133
14.6 References
134
LESSON 15
135
E2
E2 - E1 = h w
E1
15.5. Spontaneous Emission
Let us suppose that the atom is in the highest state E2 .Usually the life time of
the exited state is very small, so the atom jumps to the lower energy state E1 and it
emits the photon of frequency u (angular frequency w ) / E2 - E1 = h w . This process
is known as spontaneous emission. Its probability is determined by the properties of
the state E2 & E1
E2
E2 - E1 = h w
E1
15.6. Simulated or Induced Emission
An atom is an excited energy state E2 by the influence of electromagnetic field
of a photon can come to the lower state E1 by emitting an additional photon of same
frequency. This is known as stimulated or induced emission and its probability
proportional to the properties of the state E2 & E1 and to the spectral energy density.
E2
E 2 - E1 = h w
E1
1 N1 E1
This watermark does not appear in the registered version - http://www.clicktoconvert.com
136
The above figure represents the two energy states of an atomic system
corresponding to the energy E1 and E2 . Let N1 and N2 be the number of atoms per unit
volume in the energy level 1& 2. Atom in the lower energy level can absorb radiation
and get excited to the energy level 1 and 2. The excitation process occurs on by in the
presence of radiation. Such a process is termed as induced absorption or stimulated
absorption. The rate of absorption depend upon the energy density u( w ) associated
with the radiation field corresponding to the frequency.
E1 - E2
w =
h
The energy density u( w ) is defined in such a way that u( w ) d( w ) represent
the number of atoms per unit volume with in the interval w and w +d w . Thus the
number of absorption per unit time per unit volume is proportional to N1 u( w ).
N1 u( w ) = B12 N1 u( w )
Where B12 constant of proportionality and it is called Einsteins co-efficient
for induced absorption.
On the other hand when an atom, is an excited state, can make transition to the
lower energy state through the emission of electromagnetic radiation. However the
absorption process, emission process can occur two different ways.
The 1st refers to the spontaneous emission, in which the atom in the excited
state emits radiation even in the absence of any incident radiation. It is thus not
stimulated by any incident signal but occurs spontaneously. The rate of spontaneous
emission is proportional to the number of atoms in the excited state. If we represent
the co-efficient of proportionality by A21 then N2 A21 would represents number of
spontaneous per unit volume per unit time to the lower energy level.
The 2 nd refers to the stimulated emission in which an incident signal of
appropriate frequency triggers an atom is an excited state to emit radiation. The rate of
transition to the lower energy level is directly proportional to the energy density of the
incident radiation at the frequency w . Thus the number of stimulated emission per unit
volume per unit time is given by N2 B21 u( w ) . The quantities A21, B12, B21 are known
as Einstein co-efficients and are determined by the atomic system.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
137
N 2 A21
=
N
N 2 1 B12 - B21
N2
Or A2 1 (1)
u (w )=
N1
N B1 -2 B2 1
2
From Boltzmanns law we have the following expression for the ratio of the
population of the 2nd level at temperature T
( E2 - E1 )
N1 KB T
=e
N2
N1 hw
= e KB T
N2
where KB Boltzmanns constant. Equation (1) becomes
A2 1
u (w )= hw
KB T
e B 1 -2 B2 1
138
hw 3 1 A2 1
. hw =
2 3
p c e KB T - 1
hw
(
B2 1 e K B T - 1 )
A2 1 hw 3
=
B2 1 pc2 3
hw 3
A= B 2 3
p c
-e u r u r u r u r
Hint =
2m c
(
p. A+ A )
. p the interaction of the Hamiltonian.
e2
A 2 can be neglected because the vector potential A is very small.
2mc 2
urururur
Now we are going to prove that p. A= A ( ) ( )
. p
u r u r u r u r
In general ( p. )Ay = A.( yp)
u r u r u r u r
( p. )yA = -.ih ( yA )
u r u r u r u r
= ih-( A. )y +A. (y
)
urur
For coulomb gauge . A =0
urur
= ih- A. y
( )
uruuuuur
(
= A - hi y )
This watermark does not appear in the registered version - http://www.clicktoconvert.com
139
urur
( )
= A. py
-e u r u r-ue r u r u r u r u r u r
i.e.
2m
(
p. +A A. =p A
c 2m c
)
. +p .A p ( )
The interaction of Hamiltonian becomes
-e u r-eu r u r u r
(
= p. A . p
2A
2m c m c
) ( )
u r
Now the vector potential A can be written as
u r r r
A = $e 0 cA o-k s r(w . t )
Where e$ is the unit vector along the direction of propagation which is perpendicular
u r
to the vector potential A .
Now expressing the vector potential in terms of the exponential function
r r r r
u r ei (
k -.wr )t - i ( wk-. r )t
e+
A = e$ 0A
2
From time dependent perturbation theory the probability amplitude for transition from
an initial state n to final state s is given by
1 iws nt
cs n=s V ni n t e d t
ih
En - Es
where ws n=
h
Now,
urur
i n t n =s A . p n
s V
-e urur
= s A. p n
2m c
r r r r
-e $ A0
= e s + ei . k e r-. ip +k nr
2m c2
( - iw t
e ewi t
)
On solving the above equations the transition probability for emission become
2 w + ws n
2 2 r r
s i n t
e A $
2
2
cs n = s e p 0
n. e
i .k r
4m ch 2 2 w + ws n
2
2
This watermark does not appear in the registered version - http://www.clicktoconvert.com
140
r r
For atomic dimension the value of ei .k r
1 . So the transition probability for emission
becomes
w + ws n
2 2 s i 2 n t
e A
0 e$
2
2
cs n = s p n
4m ch
2 2 w + ws n
2
2 w + ws n
2 2 r r
s i n t
e A $
2
2
cs n = s e 0
p i .k r
n. e
4m ch
2 2 w + ws n
2
2
If we consider the atomic dimension the atomic radius is very small. So we can
expand the exponential form as
r r r r2
r r i
ei .k r= 1 .+ .
( ) (+.-i. k. )+r.
.k r
. .
1 ! 2 !
The magnitude of the successive term are
2 3
2pr 1 2pr 1 2pr
1: : :
l 2 l 6 l
r r
The magnitude of the successive term decreases by the factors of but is the
l l
r r
order of 10-3 cm. So in this case we can approximate ei .k r
=1.
2 w + w
s n
2 2 s i n t
e A
0 e$
2
2
cs n = s p n
4m ch
2 2 w + ws n
2
2
Transition for which the probability can be computed using the above
approximation is called electric dipole transition. Since only the matrix element of the
r
electric dipole moment e$ . of
r the particle is involved.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
141
The dominant factor in the nth term of the above series is proportional to the
r rn r r
( )
k . r , if k . <<1.
r Thus if the dipole matrix element vanishes, but the next term of
each series does not vanish. The transition matrix element is reduced by a factor that
has the magnitude of the order of Ka where the linear dimension of the particle wave
function of are of order of a. The transition of this type is called forbidden transition,
since a probability is reduced by a factor (Ka)2 with respect to dipole or allowed
transition.
15. 12 LET US SUM UP
In this lesson, we have studied the interaction atoms with radiation field. We
analysed spontaneous emission, induced emission , induced absorption. We
determined the Einsteins coefficients for the above three process. Also we obtained
an expression for transition probability. Finally we presented short notes on selection
rules and forbidden transition.
142
expression for the transition probability per unit time for spontaneous
emission.
5. Discuss in detail about the forbidden transition
6. Explain the selection rules.
15.14 References
1. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan. Tata
McGraw Hill Publishers.
2. Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
3. Quantum Mechanics V. Devanathan Narosa Publishing house.
4. Quantum Mechanics S.L. Gupta, V. Kumar, H.V. Sharma, R.C. Sharma.
Jai Prakash Nath & Co Ltd.
5. Quantum Mechanics Satya Prakash Sultan Chand Publishers .
6. Angular Momentum Techniques in Quantum Mechanics V. Devanathan
Kluwer Academic Publishers.
7. Quantum Mechanics John L. Powell and Bernd Crasemann Narosa
Publishing house.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
143
LESSON 16
QUANTUM FIELD THEORY
A wave field is specified by its amplitudes at all points of space and the
dependence of these amplitudes on the time, the same way as system of particulars is
specified by the positional co-ordinates qi and their dependence on the time. The field
has an infinite number of degrees of freedom and to use the amplitudes y (r,t) at all
points r as co-ordinates is analogous to a system that consists of an infinite number of
particles and with the particles co-ordinates qi(t) respectively.
y can be expanded in some complete ortho normal set of functions uk :
y (r,t) = Sk ak (t) uk (r) (1)
where ak is regarded as the fields co-ordinates In classical particle theory, both total
and partial time derivatives were defined in connection with a function F(qi, pi, t) of
the co-ordinates, momenta and time. Also a function F(y, p, t) can depend explicitly
on the time as well as on the field, so that it is important to distinguish between dF/dt
and F/dt.
144
t2
we expect the field Lagrangian to be a functional of the field amplitude y(r,t). It can
.
L = L (y , Dy ,y , t )d 3 r (2)
t2 t2 t2
3
d L dt = d L dt d r = (d L) dt d 3 r = 0 (3)
t1 t1 t1
Then summation terms in equation (6) can be integrated by parts with respect to space
co-ordinates; surface terms vanish either because y falls rapidly at infinite distance or
because y obeys periodic boundary conditions. The last term in equation (6) can be
integrated by parts with respect to time and the boundary terms vanish because of
equation (4). Now equation (6) can be written as,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
145
u
uL r r u r
t0
L L 3
- - = d yd t d r0 (7)
t1 y x y z x y
y t &
x
since equation (3) is valid for an arbitrary variation dy at each point in space,
equation (7) is equivalent to the differential equation,
u r u r u r
L L L .
- - = 0 (8)
y x yz x yy t &
x
Equation (8) is the classical field equation derived from the Lagrangian density
L (y,y,y,t).
.
u r u r
L
L Ld L L
.
= L t = -
d 0 d y d t y
xy z x y
(9)
yy r j
j j
x
Here all the dyi and dy i are zero except for particular dy j, and dti is the volume of
the cell. Similarly, the functional derivative of L w.r.t y is defined by setting all the
dy i and dy i zero except for dy i
u r
L L d L
= L t = (10)
y ddr j 0 y & jd j t& y
This watermark does not appear in the registered version - http://www.clicktoconvert.com
146
Here again at point r at which the functional derivative is evaluated is in the jth cell.
Substituting equations (9) & (10) in (8) gives,
L L
- = 0 (11)
t y& y
which closely resembles the Legrangian equations
d L L
- = 0 for a system of particle.
d t &i q i q
L
It follows that Pj = dt j (2)
y j
L
The Hamiltonian H = Pi y i - L = y i dt
. i -L (3)
i i y i
H is written as the volume integral of a Hamiltonian density H and assume that the
cells are small enough so that the difference between a volume integral and the
corresponding cell summation can be ignored; we than have,
L L
H = H d 3r H = py -L p = (4)
y y
The true field Hamiltonian H given in equation (4) which is a functional of y and p
from which y has . been eliminated.
. The classical Hamiltonian equations of motion
will be derived without further recourse to the cell a approximation. The variation of
L produced by variations of y and y can be written using equation (4) and
L / L /
- = 0
t y& / y /
L / L /
d L =d y dd+ y . p&r = d3 y d( +&p& 3
d ry )
y / y /&
This watermark does not appear in the registered version - http://www.clicktoconvert.com
147
Comparison of equations (5) & (6) for arbitrary variations dy and dp then gives the
classical field equations is Hamiltonian form.
H / H /
y& = =
p &- (8)
p / p /
The Hamiltonian equation for the time rate of change of a functional F of y and p
can now be found. We express F as volume integral of the corresponding functional
or on gradients of y or p . Hence,
u r u r
d
F F F F
= + +& p & d 3 r
y
d t t y p
F F/ H/ F/ H /
= + - d3 r
t / y / p p/ y/
F
= + {F, H} (9)
t
The above equation also serves to define poisson bracket expression for two function
as of the field variables. The RHS of equation (9) is not changed if F also depends on
y or p. It is clear from equation (9) that H is a constant of the motion if it does not
depend explicitly on time; in this case H is the total energy of the field.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
148
h2
- ih y&=* V-* y (2r +, yt ) *
2m
Which is the complex conjugate of (a) variation of y * gives
h2
ih y& = V - y 2r + y
(t , ) (3)
2m
The momentum canonically conjugate to y is
u r
L
p = =y i h * (4)
y&
However y * does not appear in the Lagrangian density so that p is identically zero. It
is therefore impossible to satisfy the corresponding classical poission bracket relation,
so that y *, p cannot be regarded as a pair of canonically conjugate variables. p
is the second momentum.
The Hamiltonian density is,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
149
u u r u r -i h -i
H = p &- yL= p y
.V p y (5)
2m h
The Hamiltonian equations of motion obtained from the classical field equations with
the help of functional derivates are,
i hi
y& = + y 2
-Vy
h
2m
i i h 2
p& = Vp- p
h 2m
We have thus shown, from the point of view of classical field theory, that the
Lagrangian density and the canonical variables and hamiltonian derived from it are in
agreement with the wave equation.
for the operations ak and ak + were solved in connection with the harmonic oscillator. It
was found that the solution of aa+- a +a=1 in the representation in which a+a i s
diagonal, consists of the matrices.
0 0 0 0
0 1 0 0 1
1 0 0 0
a = 0 0 2 2 0 a+ = (2)
0
1
2
1 0 2 0
0 0 0 3 2 1
2
0 0 3 0
Let N be Hermitian, its eigenvalues are real.
+
N= Nk
k where N k = ak ak (3)
The states of the quantized field, in the representation in which each Nk is diagonal,
are the kets
n1 , . .2. . . .n. . . . . . . .n. . . .k. , . . . . . . . . . . . . . (4)
Where each nk is an eigen value of Nk and must be a +ve integer or zero. We also
have the relations.
ak 1 , . . . n. . . . . .k . . . n. . . . 1. . , . . n.k . . =. . . 1. ./ .2 .n. . . ,k..............1,........................
. . .-. n
1 / 2 (5)
ak 1 , . . . .n. . . . .k . . . .n. . . . 1. , . . .n =k. (. . . . . .). .n.+. . 1k ., ..... . n. . +. . . . . . . . 1 , . . . . . . . . . . . . . . .
This watermark does not appear in the registered version - http://www.clicktoconvert.com
150
These ak + and ak are called creation and destruction operators for the state k of the
field.
The number operation Nk need not be a constant of the motion, that
N = Nk is a constant The rate of change of Nk is given by,
ih &Nk = ak
ak , H
h2
where H = a j la u *j V* j
ul . u+l u3 d r
j l 2m
-h 2 2
= a j l aj *
u Vl u+
3
d r (6)
j l 2m
diagonal, Nk is constant only if all the volume integrals in equation(6) are zero for
which either j or l is equal to k. These integrals are just matrix elements of the one
particle Hamiltonian. The integrals in equation (6) are than E l d jl , and the field
Hamiltonian becomes,
+
H= ak
k ak Ek = N k Ek
k
It is necessary to modify the above equations if we are obtain a theory of particles that
obey the exclusion principle. It is reasonable to require that this modification be
made in such a way that The quantum equation of motion of y is the wave equation,
h2
ih y& = V - y 2r - (yt , )
2m
when the Hamiltonian has the form,
This watermark does not appear in the registered version - http://www.clicktoconvert.com
151
h2
H = y + .y ++y
V yd 3
r
2m
It was found by Jordon and Wigner that the descried modification consists in
the replacement of the commutator brackets,
[A,B] AB BA
By anticommutator brackets
[A,B]+ AB + BA
This means that equations (1) are replaced by,
For the eigen values. These are quadratic equations that have two roots 0 and 1. Thus
the eigen values of each Nk are 0 and 1, and the particles obey the exclusion principle.
The eigen values of N = N k are the +ve integers and zero as before. The earlier
expressions for the Hamiltonian in creation, destruction and number operators are
unchanged, and the energy eigen values are n E k k .
This watermark does not appear in the registered version - http://www.clicktoconvert.com
152
16.11 References
1. A text book of Quantum Mechanics P.M. Mathews and K. Venkatesan. Tata
McGraw Hill Publishers.
2. Quantum Mechanics Leonard I. Schiff, McGraw Hill International edition.
3. Quantum Mechanics V. Devanathan Narosa Publishing house.
4. Quantum Mechanics S.L. Gupta, V. Kumar, H.V. Sharma, R.C. Sharma.
Jai Prakash Nath & Co Ltd.
5. Quantum Mechanics Satya Prakash Sultan Chand Publishers .
6. Angular Momentum Techniques in Quantum Mechanics V. Devanathan
Kluwer Academic Publishers.
7. Quantum Mechanics John L. Powell and Bernd Crasemann Narosa
Publishing house.
This watermark does not appear in the registered version - http://www.clicktoconvert.com
153
154
155